Book — Chapter Two Tone

Chapter Two: Strict and Broad Ideas and the Scope of Objective Reality

Introduction

In the Third Meditation, Descartes states that the objective mode of being belongs to ideas by their own nature (AT VII 42; CSM II 29).¹My translation, “by their own nature,” is philosophically preferable to Cottingham’s “by their very nature” as a translation of Descartes’s Latin: “iste modus essendi objectivus competit ideis ex ipsarum naturâ.” (AT VII 42; CSM II 29)

The crucial expression is ex ipsarum naturâ. The emphatic pronoun ipsarum directs attention to the ideas themselves, whose nature grounds the objective mode of being. The preposition ex identifies that nature as the source or ground of the objective mode of being.

There are three reasons to prefer “own”:

(1) Grammatical precision. “Their own nature” makes the reference to the ideas themselves explicit. “Their very nature” is grammatically defensible, but it emphasizes the nature itself rather than its belonging to the ideas.

(2) Philosophical precision. “Own” directs attention to what belongs to the ideas under discussion in virtue of their nature; I identify those ideas as strict ideas from the surrounding argument. It does not, by itself, establish that every mental mode qualifies as an idea possessing objective reality.

(3) Interpretive significance. The universalist interpretation cannot derive its unrestricted scope merely from ipsarum. It must independently establish which cognitions fall within the extension of ideis in this passage.

There is, however, one qualification. The Latin ipsarum can legitimately be translated as “very” when “very” functions as an intensifier of identity rather than degree. My translation makes the reference to the ideas themselves more explicit, but neither rendering determines whether Descartes means every thought he calls an idea or only the strict ideas whose represented contents he has been comparing. Which ideas? If he means every thought he calls an idea, the statement supports the stronger thesis: sensations, appetites, volitions, and every other broad-sense idea must contain objectively real representational content. Yet the statement occurs within an argument comparing ideas according to the things they represent. I argue that Descartes is speaking exclusively about strict ideas here and that his explanation of their nature does not establish what every conscious thought must contain.

A strict idea internally represents a subject matter through objectively real content. The representatum, or item represented, has objective being in the idea, and its ontological rank determines the content’s degree of objective reality. Descartes’s broad use also includes sensations, willing, fearing, and other immediately experienced thoughts. The weaker thesis requires every one of these thoughts to represent something internally; the stronger thesis requires each to contain objectively real content. To defend either universal thesis, an interpreter must explain why the relevant requirement applies to sensations and acts of willing as well as to strict ideas. Descartes’s use of the same name for all these thoughts does not establish that they contain the same kind of content.

We must therefore determine whether Descartes’s statements attributing objective reality to ideas apply to every Cartesian idea or only to ideas understood properly and strictly as internally representing things. Descartes’s classification of thoughts in the Third Meditation and his subsequent comparison of what different ideas represent provide grounds for restricting the scope of those statements. The reply reported in the Conversation with Burman distinguishes ideas taken “properly and strictly” from the broader use for common notions, providing further evidence that apparently unrestricted statements about ideas of things apply exclusively to strict ideas. (AT V 153; CSMK III 338) Refuting the universal claims also requires examples of complete ideas that lack the content attributed to them. Chapters Four and Five supply those arguments by examining sensations that lack objective reality and by establishing that bare secondary-quality sensations lack internal representation.

2.1. What Qualifies as a Cartesian Idea?

In the geometrical exposition appended to the Second Replies, Descartes defines “idea” as follows.

I understand this term [‘idea’] to mean the form of any given thought, immediate perception of which makes me aware of the thought. (AT VII 160; CSM II 113)²Second Replies, Definition II. The definition continues across AT VII 160–61. English quotations follow CSM unless another translation is identified.

Under this definition, a thought qualifies as an idea because the thinker is immediately aware of it. Descartes does not add that something must be objectively represented within it. Anyone who attributes represented content to every thought covered by “any given thought” must explain why sensing, willing, fearing, and the other conscious activities require it.

Continuing Definition II, Descartes distinguishes the form of a thought from a corporeal image depicted in the brain. The brain image itself does not count as the mental idea. Imaginative thoughts qualify because the mind is immediately aware of them, and the definition extends to other thoughts on the same basis. (AT VII 160–61; CSM II 113) Excluding a bodily picture from the definition leaves open what the conscious thought contains.

When Hobbes challenges his use of “idea,” Descartes explicitly includes wanting and fearing. In the Third Replies, he takes the term to refer to whatever the mind immediately perceives and explains why these activities qualify.

For example, when I want something, or am afraid of something, I simultaneously perceive that I want, or am afraid; and this is why I count volition and fear among my ideas. (AT VII 181; CSM II 127)

Wanting and fearing qualify because the person immediately perceives these activities. Descartes thus includes the volitional or affective thought itself within the broad use of “idea.” He does not reserve the term for the representation of the object while excluding the experienced wanting or fearing.

Being aware of willing does not require a second act directed at the willing. In the Letter to Mersenne of 28 January 1641 (AT III 295; CSMK III 172), Descartes denies that the idea by which willing is known must differ from the willing itself. The person is aware of willing in the act of willing.³ A later reflective thought takes that act of willing as its represented subject matter. Such reflection may contain objectively real content representing the earlier act. But this does not show that the original act of willing represented itself or that its volitional aspect consisted in objectively real representational content.

Since these examples involve wanting something and fearing something, neither establishes that the complete thought lacks objective reality. Fearing a lion includes objectively real content representing the lion. The passage establishes that the experienced fear also belongs to the thought Descartes calls an idea, even though the fear and the represented lion perform different roles within it.

Sensations also fall within this broader use. In the Fourth Replies (AT VII 234; CSM II 163), Descartes calls color and cold ideas arriving from the senses while considering their failure to represent anything real. Their disputed objective reality gives no reason to exclude them from his broad use of “idea.”⁴David Cunning excludes sensations from the category of ideas: “A Cartesian sensation is a mode of mind but not an idea. If it is representational, it is not representational in virtue of having objective reality but in virtue of something else.” (“Descartes on Sensations and Ideas of Sensations,” in An Anthology of Philosophical Studies, vol. I, edited by Patricia Hanna, Adrianne Leigh McEvoy, and Penelope Voutsina, Athens, Greece: Atiner Publishing, 2006, 18)

Cunning later explains his restriction: “Another is that ideas are representational. Indeed, Descartes says that although something can be dubbed a ‘thought’ in a loose sense so long as it is a mode of mind, the term ‘idea’ is to be reserved for modes of mind that are of objects.” (23)

This restriction fits the strict use of “idea” in the Third Meditation (AT VII 37; CSM II 25–26), but Descartes also recognizes a broader use. In the Fourth Replies, he expressly calls colour and cold sensory ideas while considering their failure to represent anything real (AT VII 234; CSM II 163).

I agree with Cunning that secondary-quality sensations lack objective reality, but reject his exclusion of sensations from Cartesian ideas.

Cunning also locates material falsity in an idea of a sensation (18), whereas I argue that a materially false sensation itself has a misleading presentational character before any false judgment is made. His exclusion of sensations therefore differs from my rejection of the universal theses, which concern every idea in Descartes’s broad sense. The dispute is over what these acknowledged sensory ideas contain.

Calling a thought an operation of the intellect does not yet establish that it contains objectively real representational content, because Descartes uses “intellect” more broadly than “pure understanding.” In Principles of Philosophy I.9 (AT VIIIA 7; CSM I 195), he includes understanding, willing, imagining, and sensory awareness among the activities of thought. In I.32, he assigns perception to the intellect and volition to the will, then states that “Sensory perception, imagination and pure understanding are simply various modes of perception.” (AT VIIIA 17; CSM I 204) Sensation therefore belongs to the intellect when “intellect” means the faculty of perception, although it is not an act of pure understanding. Calling a cold sensation an activity of the intellect tells us that Descartes uses “intellect” broadly enough to include sensing. That classification does not establish that the felt coolness is internally represented through objectively real content. Feeling coolness and internally representing something remain distinct, even though Descartes includes both sensing and pure understanding among the intellect’s activities.

Feeling fear or sensing coolness is different from internally representing something, as the comparison between fearing a lion and feeling cold illustrates. In fearing a lion, the fear is immediately experienced, while the lion is internally represented through objectively real content. Both belong to one thought, but the fear is a formally real affective determination of the mental mode rather than another internally represented item alongside the lion. A cold sensation likewise has an immediately experienced character: its phenomenal coolness is a formally real qualitative determination of the sensory mode. On my interpretation, however, the bare sensation contains no internal representatum. I defend the absence of objectively real content in Chapter Four and the absence of internal representation in Chapter Five. The coolness is phenomenally presented rather than objectively represented. Descartes’s broad definition counts both fearing a lion and sensing coolness as ideas because both are immediately experienced. The lion is internally represented in the fear of a lion via an objectively real representstion, but a cold sensation need not contain a comparable representation to qualify as an idea. Denying that the sensation contains objectively real content therefore gives no reason to deny that the sensation is a Cartesian idea.

2.2. Strict Ideas and the Image Analogy

In the Third Meditation, Descartes distinguishes different kinds of thought to determine which can properly be true or false.

Some of my thoughts are as it were the images of things, and it is only in these cases that the term ‘idea’ is strictly appropriate—for example, when I think of a man, or a chimera, or the sky, or an angel, or God. (AT VII 37; CSM II 25)

The Latin description is tanquam rerum imagines, which I render as “as if images of things.” The expressions quædam, “some,” and quibus solis proprie convenit ideæ nomen, “to which alone the name ‘idea’ properly belongs,” identify the thoughts receiving the strict name. Descartes then describes willing, fearing, affirming, and denying as having additional forms. I take their volitional, affective, and judgmental character to be distinguishable from the represented subject matter within the same thought.

Because Descartes includes an angel and God among his examples, “as if images” cannot require a visual picture. A strict idea specifies what is thought of, its representatum, even when that subject matter is understood without a visual likeness. Nor must a corresponding thing exist outside the thought.

A chimera, with a lion’s body, a serpent for a tail, and a goat’s neck and head projecting from its back, illustrates why a represented thing’s nonexistence does not deprive its idea of objective reality. No such creature exists, yet Descartes counts its idea among strict ideas. The represented creature has objective being in the thought without existing formally, i.e., actually. Consequently, an idea of a nonexistent creature does not refute the universal attribution of objective reality. A counterexample must be an idea that lacks objectively real representational content, whereas the chimera is objectively present in the idea despite having no formal existence outside it.

In Definition III of the Second Replies, Descartes explains objective reality by reference to represented content.

[…] the being of the thing which is represented by an idea, in so far as this exists in the idea. (AT VII 161; CSM II 113)⁵Second Replies, Definition III. The initial ellipsis omits the introductory words of the definition.

Applied to an idea of a lion, Definition III identifies the lion’s objective being with its being represented in the idea. I understand objective reality as the internally contained representational content through which the representatum is present in the intellect. Its degree depends on the representatum’s ontological rank as a mode, a finite substance, or an infinite substance. An idea of a lion represents a finite substance, while the idea of God represents an infinite substance and therefore, on Descartes’s account, contains infinite objective reality.

Does a cold sensation contain a represented thing in the way that an idea of a lion does? Calling both thoughts ideas does not answer that question. Definition III explains objective reality through what exists in an idea as represented, so extending its account to every broad-sense idea requires identifying represented content in each case.

Both uses of “idea” are deliberate. Descartes reserves the strict name for thoughts that are as if images of things (tanquam rerum imagines), gives the broad definition through immediate awareness in the Second Replies, and defends that broader use against Hobbes by including volition and fear.⁶The Third Meditation identifies thoughts that receive the name “idea” properly and strictly through their represented subject matter. (AT VII 37; CSM II 25) Definition II in the Second Replies explains an idea through immediate awareness of a thought. (AT VII 160–161; CSM II 113–114) The reply to Hobbes expressly includes volition and fear because they are immediately perceived. (AT VII 181; CSM II 127–128) The Fifth Replies also rejects restricting ideas to images in the imagination. (AT VII 366–367; CSM II 253)

The inclusion of volition and fear establishes that the broad use covers the thought with its experienced volitional or affective character. It does not establish that the complete thought lacks represented subject matter. Fearing a lion includes objectively real content representing the lion, although the experienced fear is not itself another represented item. Nadler can therefore accept these passages while maintaining that every broad-sense idea contains strict representational content. The distinction between the two uses must be preserved, but neither the presence nor the absence of objective reality in every thought admitted under the broad use follows from that distinction alone. Descartes calls a thought an idea in the strict sense because it internally represents a thing. His broad definition counts a thought as an idea because the thought is immediately experienced. Calling a sensation an idea in this broader sense therefore does not establish that the sensation internally represents something. The requirement of internal representation belongs to the strict use and cannot simply be attributed to every thought Descartes calls an idea.

2.3. Fearing a Lion and Having a Representational Aspect

Willing and fearing make it necessary to distinguish the thought’s represented subject matter from its volitional or affective character. Such thoughts can internally represent something while also containing features that are not themselves represented. Fearing a lion, for example, involves objectively real content representing the lion together with the formally real experience of fear. Must the affective character also count as objectively real representational content merely because it belongs to the same thought? I argue that it need not. The presence of objective reality within a thought does not make every feature of that thought internally representational. Descartes introduces willing and fearing immediately after his discussion of ideas understood properly and strictly:

Other thoughts have various additional forms: thus when I will, or am afraid, or affirm, or deny, there is always a particular thing which I take as the object of my thought, but my thought includes something more than the likeness of that thing. Some thoughts in this category are called volitions or emotions, while others are called judgements. (AT VII 37; CSM II 26)

Merely thinking of a lion need not involve fearing it. In fear, the thought contains objectively real content representing the lion together with an experienced affective determination. The fear changes how the person experiences the thought without itself becoming another representatum.

Although the representation of the lion and the experienced fear are distinguishable, they belong to one mental mode. Descartes’s contrast between the likeness of the thing and what the thought includes beyond that likeness supports this distinction within the thought. It gives no reason to turn the representation and the fear into independently occurring acts.

A universalist can nevertheless appeal to Descartes’s word semper, “always.” In this passage he assigns an apprehended object to willing, fearing, affirming, and denying, and fear of a lion fits that description. To extend the account to every conscious thought, however, the interpreter must show that a bare cold sensation, objectless dread, and the idea of nothing likewise contain represented subject matter.

For a cold sensation, what would take the place of the lion? Felt coolness is present in the experience, but a commentator who extends the account of fearing a lion to this sensation must identify what it internally represents and explain why the feeling requires that representation. Chapter Four argues that secondary-quality sensations lack objectively real content. Chapter Five argues that they also lack internal representation and explains how their lawful connections with bodily causes allow them to function as natural signs.

Focusing on the non-representational fear within fear of a lion would therefore leave the universalist’s claim intact. The complete thought still contains objectively real content representing the lion. A successful counterexample must lack that content throughout the idea, rather than merely have an affective feature that is not itself represented.

With objectless dread, the whole experience supplies the counterexample that fear of a lion cannot provide. Its felt character constitutes a complete affective experience without objectively real content representing a lion, danger, or another subject matter. Descartes need not have discussed it for it to qualify under his broad definition: the person immediately experiences this mental state. In Chapter Five, I defend its absence of objective reality and internal representation. An affective state of this kind disproves both universal theses. Descartes’s description of fear directed at an object cannot establish that every affective experience has such an object.

Steven Nadler nevertheless extends the structure illustrated by fearing a lion to every broad-sense idea: “At the core of every idea in the broad sense is an idea in the strict sense, giving it a specific content or referent.”⁷Steven Nadler, “The Doctrine of Ideas,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 87. In fearing a lion, an individual thought contains objectively real content representing a lion together with the formally real non-internally representational fear. Nadler claims that every idea in the broad sense likewise contains a strict idea that supplies its internally represented subject matter. Applying his claim to a cold sensation requires identifying something internally represented within the sensation, just as the lion is represented in fearing it. Nadler also treats objective reality as what gives ideas their represented subject matter, making Descartes’s explanation of objective reality directly relevant to his interpretation. In the Third Meditation, Descartes compares ideas according to what they represent, distinguishing the objective reality of ideas of modes, finite substances, and God. In Section 2.4, I argue that Descartes’s comparisons presupposes represented subject matter in the ideas being compared. If Nadler uses Descartes’s comparison of degrees of objective reality to argue that a cold sensation internally represents something, he must first include secondary quality sensations among ideas with represented subject matter. For example, if he assumes that a sensation represents a bodily mode, Descartes’s ranking assigns its content the degree of objective reality appropriate to a mode. That ranking explains the degree of objective reality the supposed representation would have. It does not establish that the sensation contains the representation attributed to it. I dispute precisely that attribution. Unless Nadler supplies an independent reason for accepting it, his appeal to Descartes’s ranking assumes that the sensation contains internally represented content and therefore begs the question.

2.4. Ideas Considered as Thoughts and as Representing Things

Before comparing ideas according to their objective reality, Descartes examines why he has taken some of them to originate outside himself. A natural impulse and the apparent independence of sensory ideas from his will do not establish their origin or resemblance to external things.⁸Descartes distinguishes “a spontaneous impulse” to believe that ideas resemble external things from knowledge supplied by the natural light. He recalls that natural impulses have previously misled him. (AT VII 38–39; CSM II 26–27)

He also allows that a faculty within himself, not yet fully understood, could produce ideas independently of his will without assistance from external things. Finally, even an external origin would not guarantee resemblance. His sensory and astronomical ideas of the sun differ, and he rejects the sensory appearance as an accurate likeness of the sun. (AT VII 39; CSM II 27) He then announces a different approach:

[…] there is another way of investigating whether some of the things of which I possess ideas exist outside me. (AT VII 40; CSM II 27)

Descartes first considers ideas “quatenus ideae istae cogitandi quidam modi tantùm sunt, non agnosco ullam inter ipsas inaequalitatem,” translated by Cottingham as “In so far as the ideas are simply modes of thought, there is no recognizable inequality among them.” (AT VII 40; CSM II 27) He then considers them “sed, quatenus una unam rem, alia aliam repraesentat, patet easdem esse ab invicem valde diversas.” [But in so far as different ideas represent different things, it is clear that they differ widely.] (AT VII 40; CSM II 28)

In both clauses, quatenus, “insofar as,” specifies how the ideas are being considered. As modifications of a finite mind, they have the same formal rank. Their differences in objective reality emerge when Descartes considers what they represent. Ideas of substances contain more objective reality than ideas of modes or accidents, and the idea of an infinite substance contains more than ideas of finite substances. These differences in represented reality determine how much reality their adequate causes must possess.

Thinking of a lion illustrates both ways of considering an idea. As a mental mode, the idea exists formally; as representing a finite substance, it contains the corresponding objective reality. The lion has objective being as represented without existing in the mind as a physical animal. We can distinguish the mental mode from its represented content while recognizing that both belong to this one idea.

With the idea of God, the difference between the act’s formal reality and its content’s objective reality is especially clear. A finite modification of mind represents an infinite substance. Descartes argues that only a cause with infinite reality can account for that objectively real content. (AT VII 40–42; CSM II 28–29)

These degrees measure the ontological rank of what is represented. Making an idea more vivid, intense, or complex does not, by itself, change that rank.

Nothing in the two uses of quatenus alone establishes that some ideas lack objective reality. A universalist could accept both ways of considering an idea and maintain that both apply to every conscious thought. Descartes introduces objective reality here by comparing ideas through the different things they represent. The second clause concerns ideas that represent one thing rather than another; it should not be generalized into the separate thesis that every broad-sense idea necessarily has an objectively real representational aspect.

Why, then, does an idea contain the particular objective reality it does? Its cause must possess at least as much reality, formally or eminently, as the idea contains objectively. (AT VII 41; CSM II 28–29) Accounting for the occurrence of a finite mental mode therefore does not exhaust the causal question. The cause must also account for what the idea represents.

In the “Preface” to the Meditations, Descartes distinguishes an idea taken materially, as an operation of the intellect, from an idea taken objectively, as the thing represented by that operation. (AT VII 8; CSM II 7) This distinction differs from the broad and strict uses of “idea.” A strict idea is an operation of the intellect and can be considered both materially and objectively. Describing a cold sensation as an operation of the intellect establishes its status as a mental mode; whether it also contains a represented thing by way of objectively real content is the disputed claim.

We must also distinguish an idea’s formal reality from what Descartes calls taking an idea formally in the Fourth Replies. He writes that “when we think of them as representing something we are taking them not materially but formally.” (AT VII 232; CSM II 163) To take the idea of a lion formally in this passage is to consider it as representing the lion. Its formal reality, by contrast, consists in its actual existence as a modification of mind. The terminology must not obscure the difference between the thought’s occurrence and its representation of the lion through objectively real content.

It is while explaining represented content and its causes that Descartes states that objective being belongs to ideas by their nature. I read this as a claim about what strict ideas necessarily possess. The next section follows the argument around that sentence to show why “ideas” there refers to thoughts with represented subject matter, such as the ideas of a lion and of God.

2.5. Why Objective Being Belongs to Strict Ideas by ‘Their Own Nature’

Why does the objective mode of being belong to ideas by their own nature? Descartes answers while explaining why their causes must possess sufficient reality formally or eminently. I translate ex ipsarum naturâ as “by their own nature,” preserving the reference to the ideas themselves. Which ideas he means must be established from the argument. The sentence reads:

Nam quemadmodum iste modus essendi objectivus competit ideis ex ipsarum naturâ, ita modus essendi formalis competit idearum causis, saltem primis & præcipuis, ex earum naturâ. [My preferred translation: For just as this objective mode of being belongs to ideas by their own nature, so the formal mode of being belongs to the causes of ideas—at least the first and principal causes—by their nature.]⁹I prefer “by their own nature” for ex ipsarum naturâ because ipsarum emphasizes the ideas themselves. The feminine plural pronoun refers to the ideas mentioned by ideis; its genitive case identifies whose nature Descartes means. Cottingham’s “by their very nature” is defensible, but “their own” makes this reference more direct. Both translations preserve the necessity of what belongs to the ideas by nature.

The concluding ex earum naturâ, “by their nature,” refers to the causes of ideas. Descartes uses the ordinary earum here, whereas the earlier ipsarum is emphatic. Cottingham renders both phrases with “their very nature.” My translation retains the difference.

I translate saltem primis & præcipuis as “at least the first and principal causes.” Primis identifies the first causes; præcipuis identifies the principal or chief causes. Descartes immediately explains why the causal account must reach them: an indefinite succession of ideas originating from other ideas would leave unexplained the source of the reality they contain objectively. The ultimate cause must possess sufficient reality formally or eminently. “Principal” makes this explanatory role clearer than Cottingham’s “most important,” although his wording is lexically defensible. Repeating “causes” also keeps that role explicit.

The idea of God illustrates the issue. Its occurrence is a finite mental mode, but its content represents an infinite substance. The cause must account for that infinite objective reality. The same requirement applies according to the appropriate rank in other cases: an idea of a lion contains the objective reality of a finite substance.

Finally, iste, “this,” points back to the objective mode of being just explained. Cottingham’s “the” leaves that reference implicit. Translating it as “this” makes the continuity of the argument more apparent, but the demonstrative does not by itself restrict the statement to strict ideas. Nor does ipsarum supply that restriction merely by emphasizing the ideas. My identification of them as strict ideas rests on Descartes’s comparison of represented contents and his explanation of the reality their causes must possess. (AT VII 42; CSM II 29)

[Cottingham’s translation: For just as the objective mode of being belongs to ideas by their very nature, so the formal mode of being belongs to the causes of ideas—or at least the first and most important ones—by their very nature. (AT VII 42; CSM II 29)]

A strict idea cannot represent a lion without the lion having objective being in it. On my interpretation, its objectively real content accounts for that representation. Descartes’s appeal to the ideas’ nature expresses the necessity of this connection.

Although Descartes writes ideis, “to ideas,” without adding “strict” or “broad,” the preceding argument identifies which ideas he means. Descartes has compared them according to what they represent and asked why their different degrees of objective reality require adequate causes. His statement explains how the represented reality exists in those ideas and why its ultimate source cannot be merely another representation.

To see which ideas the argument concerns, begin with Descartes’s examples of thoughts to which the strict name belongs, including ideas of a man, a chimera, an angel, and God. (AT VII 37; CSM II 25) Each has represented subject matter, even when no corresponding thing exists outside thought. Descartes can therefore compare their objective contents before establishing the existence of anything beyond the mind. The chimera illustrates why the comparison remains possible: although no such creature exists, its idea still contains represented subject matter.

When Descartes distinguishes ideas as modes of thought from ideas considered according to what they represent, the comparison becomes one of ontological rank. Ideas of finite substances contain more objective reality than ideas of modes, and the idea of an infinite substance contains more than ideas of finite substances. (AT VII 40; CSM II 27–28) Immediate awareness alone gives none of these differences. They arise from what the ideas represent and determine how much reality an adequate cause must possess.

Once ideas have been compared in this way, their different degrees of objective reality require a causal explanation. Why does an idea contain this particular objective reality? Descartes requires its cause to possess at least as much reality formally or eminently as the idea contains objectively, explaining that objective being is not nothing and cannot come from nothing. (AT VII 41; CSM II 28–29) The reality for which he is seeking a cause is represented content. This does not turn every real feature of a thought into an objectively represented item. In fearing a lion, the fear is formally real as an affective determination of the mental mode, but its reality does not make it another represented object.

Descartes then refers to iste modus essendi objectivus, “this objective mode of being” (AT VII 42; CSM II 29), returning to the way a represented thing exists in the intellect through an idea. The comparison quemadmodum … ita, “just as … so,” connects that objective being with the formal being belonging to the causes of ideas. The represented lion has objective being in the idea, and whatever ultimately accounts for that content must possess sufficient reality formally or eminently. The sentence continues the explanation of represented reality and its cause.

Referring back to objective being with iste, “this,” does not by itself restrict the claim to strict ideas. A universalist can accept the reference and still argue that every broad-sense idea contains represented content. The restriction follows from the work Descartes is doing throughout the argument. He identifies thoughts that represent things, compares their contents, and explains what can cause them; he has supplied no argument here that every sensation or appetite contains such a representation.

Immediately afterward, Descartes again calls ideas images, using veluti quasdam imagines, “like certain images” (AT VII 42; CSM II 29). The reality they contain cannot exceed that of their ultimate sources. An idea of a stone, for example, contains objectively real content representing a corporeal substance, and its cause must possess sufficient reality to account for that content. The renewed image analogy continues the discussion of ideas that represent things. It gives no indication that Descartes has shifted to every thought included under his broad definition.

On either side of the statement about ideas’ nature, Descartes is discussing thoughts that represent things. Before it, he compares their represented contents; after it, he again describes ideas as images. The sentence itself explains the causes required for the reality contained objectively in those ideas. I therefore read it as characterizing strict ideas, with support from the surrounding argument independently of the counterexamples developed later in the book.

A universalist may accept this account of the passage and add that every broad-sense idea contains a strict representational aspect. Fear of a lion shows how that can happen. Descartes can consider the represented lion when comparing objective reality without counting the fear as another representation. But what plays the role of the lion in a cold sensation or in the idea of nothing? The explanation of fear supplies no answer for those cases.

A defender of the weaker thesis may hold that every broad-sense idea internally represents something while allowing that some lack objective reality. This offers a way to preserve universal internal representation after rejecting universal objective reality. A passage explaining the objective being of represented contents cannot itself establish that representation survives without such content. In Chapter Five, I argue against that separation by examining what secondary-quality sensations are alleged to represent and how the proposed representation could occur.

Restricting the statement to strict ideas preserves the necessity Descartes attributes to their objective being. Every strict idea contains objectively real representational content, while a thought lacking that content can still qualify as an idea under the broad definition through immediate awareness. Descartes makes this broader use explicit in the Third Replies, where he explains to Hobbes that volition and fear count among his ideas because he immediately perceives himself willing or fearing (AT VII 181; CSM II 127). These examples do not establish that the complete thoughts lack objective reality, since fearing a lion also includes represented content concerning the lion. They establish why the experienced willing or fearing belongs within his broad use of “idea.” A thought’s failure to qualify as a strict idea therefore leaves its status as a broad-sense idea intact, provided it meets Descartes’s requirement of immediate awareness.

Descartes’s causal proof concerns the idea of God and the cause of its infinite objective reality. The proof does not also require a sensation of cold to contain objectively real content. But this leaves open what Descartes believed about the sensation. He could have thought that every idea contains objective reality even though his proof did not require that universal claim. My reason for restricting his statement at AT VII 42 is therefore the argument surrounding it, where he compares ideas according to what they represent and asks what could cause their represented content. To establish that a cold sensation lacks objective reality, I must separately examine what the sensation contains.

My Restriction Thesis holds that strict ideas necessarily contain objectively real representational content, while some broad-sense ideas contain none. An idea of a lion illustrates the first part. It represents the lion through objectively real content, and without that content it would not be a strict idea of a lion. A bare sensation of cold illustrates the second part on the interpretation I defend. Its felt coolness makes it a conscious experience, but experiencing coolness does not itself establish that the sensation internally represents anything. In Chapter Four, I examine what the sensation is alleged to represent and argue that it contains no objectively real content. Chapter Five then addresses the further claim that it might internally represent something even without objective reality. There I argue that bare secondary-quality sensations lack even internal representation and explain how they nevertheless signify their bodily causes. The reading defended in this section establishes why Descartes’s statement about the nature of strict ideas does not settle those questions in the universalist’s favor.

A defender of universal internal representation can nevertheless appeal to Descartes’s statement in the discussion of heat and cold, “And because there can be no ideas except as if of things” (AT VII 44; CSM II 30, my translation). The wording “no ideas except” excludes exceptions, and its occurrence in a discussion of sensations gives the universalist a direct textual basis for claiming that even a cold sensation must represent something. The reply reported in the Conversation with Burman answers the question of scope by specifying that “idea” in this statement is understood properly and strictly, while common notions belong to a broader use (AT V 153; CSMK III 338). The statement therefore governs strict ideas. Extending its requirement to every conscious thought would substitute the broad sense of “idea” for the strict sense expressly identified in the reply. Calling cold a sensory idea does not justify that substitution or establish what the sensation internally represents. Section 2.6 develops this answer to the universalist reading and examines how the restriction bears on Descartes’s discussion of sensations of heat and cold.

2.6. Apparently Universal Language and the Reply to Burman

While considering the obscure sensory ideas of heat and cold, Descartes makes a statement that a defender of universal internal representation can invoke against my interpretation. Its wording excludes exceptions, so the question is which ideas it includes:

Et quia nullæ ideæ nisi tanquam rerum esse possunt. [Cottingham’s translation: “And since there can be no ideas which are not as it were of things. . .”] [My translation: And because there can be no ideas except as if of things.] (AT VII 44; CSM II 30)¹⁰My translation of Et quia nullæ ideæ nisi tanquam rerum esse possunt is “And because there can be no ideas except as if of things.” (AT VII 44; CSM II 30) I prefer “as if” to Cottingham’s “as it were” because the choice affects how the reader understands an idea’s being of something. “As it were” readily invites a qualified attribution, whereas “as if” foregrounds the difference between how something appears or behaves and whether it possesses the attributed character. Consider two descriptions of someone’s position in a department. Saying that he is, as it were, the department’s father credits him with a figurative paternal role. Saying that he behaves as if he were the department’s father describes his conduct without granting him that role. He might simply be assuming an authority he lacks. The distinction is between attributing a character in a qualified sense and describing something as if it possessed that character without making the attribution.

Descartes’s discussion of bodily pain supplies a directly relevant example. In the Sixth Meditation, he writes “sentietur dolor tanquam in pede,” which Cottingham translates as “then pain will be felt as if it were in the foot.” (AT VII 88; CSM II 61) Cottingham here renders tanquam as “as if.” The surrounding explanation concerns a bodily disturbance occurring elsewhere along the nerves, or in the brain itself, that produces the same brain motion ordinarily caused by injury to the foot. The resulting pain is nevertheless felt in the foot. Its felt location belongs to the sensory experience itself. The person need not first experience an unlocated pain and then judge that it is in the foot. “As if” expresses this bodily location as experienced without affirming that the bodily disturbance occurs there.

A cold sensation requires the same attention to what is already present in the experience. On my interpretation, phenomenal coolness is a formally real qualitative determination of the sensory mental mode, and it can be felt in the left hand rather than the right. Its being a mental determination does not prevent it from being experienced as bodily located. The sensation can therefore be of cold through its felt coolness and can present that coolness as located in the hand before any judgment is made. But coolness felt in the hand, the ice cube’s causing that experience, and the attribution of the same phenomenal coolness to the ice are different matters. Neither the felt bodily location nor the identification of the ice as the cause establishes that the ice possesses the qualitative character experienced. Nor should the sensory attribution of location simply be identified with an objectively real representation of the hand. Chapter Five examines the argument that felt bodily location requires such an internal representation.

These examples explain the philosophical importance of translating tanquam rerum as “as if of things.” The qualification permits the experience’s own apparent bodily attribution to be acknowledged without treating that attribution as an established account of what the body possesses or what the sensation internally represents. By contrast, “as it were of things” can encourage a reader to assume that a thing is already represented and to qualify only the manner of its representation. For example, the sensation may be taken to contain an obscure representation of coolness in the ice, although the presence of that represented bodily quality is precisely what requires argument. My translation makes this step visible. Just as pain felt as if in the foot does not establish that its bodily cause is in the foot, coolness experienced as if belonging to a body does not establish that the body possesses phenomenal coolness. The force of tanquam is to keep the sensory appearance explicit, so that what is felt is not silently converted into an assertion about the body or an assumption about the sensation’s objectively real content.

Burman challenges this requirement with the idea of nothing. In Cottingham’s translation of the reported exchange, his objection reads, “But we have an idea of nothing, and this is not an idea of a thing.” (AT V 153; CSMK III 338) If the requirement covers everything Descartes calls an idea, this example contradicts it. Descartes’s reply identifies the sense of “idea” in which the requirement holds:

Illa idea est solùm negativa, et vix vocari potest idea; auctor autem hîc sumit ideam proprie et stricte. Aliæ etiam dantur ideæ notionum communium, quæ non sunt ideæ rerum proprie; sed tum idea latius sumitur. [Cottingham’s translation: {Descartes} That idea is purely negative, and can hardly be called an idea. In this passage the author is taking the word ‘idea’ in its strict and narrow sense. We do also have ideas of common notions, which are not, strictly speaking, ideas of things. But this is a rather extended use of the word ‘idea’.]

[My translation: That idea is only negative and can scarcely be called an idea; the author, however, takes “idea” here properly and strictly. There are also ideas of common notions, which are not properly ideas of things; but then “idea” is taken more broadly.] (AT V 153; CSMK III 338)¹¹My translation of the reported reply in the Conversation with Burman, 16 April 1648. This is a report of the exchange, rather than a passage from a work published by Descartes. The AT apparatus identifies Burman’s objection as directed at the statement that ideas can only be as if of things (AT VII 44; CSM II 30). Burman is not directly challenging the earlier statement that objective being belongs to ideas by their nature (AT VII 42; CSM II 29).

Descartes’s reply therefore expressly restricts the statement about ideas being as if of things to strict ideas. It also supports my reading of the earlier statement by showing that Descartes’s apparently universal language about ideas does not automatically include every thought he calls an idea. The argument for restricting AT VII 42 comes from his comparison of represented contents and their causes, as developed in Section 2.5.

My translation preserves several features of the Latin that explain how Descartes’s reported reply to Burman restricts the claim that ideas can only be as if of things. Burman challenges that claim with the idea of nothing. If every idea must be of a thing, an idea of nothing appears to be a counterexample. Descartes’s reply identifies the relevant use of “idea” as proprie et stricte, “properly and strictly,” and then acknowledges ideas of common notions that non sunt ideæ rerum proprie, “are not properly ideas of things.” He explains their inclusion by adding that idea latius sumitur, “‘idea’ is taken more broadly” (AT V 153; CSMK III 338). By distinguishing these uses of “idea,” Descartes limits the requirement to strict ideas while expressly recognizing common notions as ideas in the broader sense. Someone defending universal internal representation cannot therefore assume that every thought Descartes calls an idea must satisfy the requirement he attributes to ideas properly and strictly understood.

I translate proprie et stricte as “properly and strictly,” preserving both terms. Cottingham’s “in its strict and narrow sense” emphasizes that fewer thoughts fall within this use of “idea,” but “properly” directs attention to what qualifies a thought for that use. An idea of a lion is a strict idea because it internally represents a lion. The question is whether every other thought called an idea must likewise contain an internal representation of a thing. Descartes’s reply expressly distinguishes common notions from ideas of things properly so called. Understanding a principle such as nothing comes from nothing therefore cannot simply be treated as another case of representing a thing in the way that the idea of a lion does. The principle qualifies as an idea under the broader use, while its actual content must be examined to determine whether understanding it involves any represented thing. Preserving “properly” makes the reason for the distinction clearer. The broader category includes thoughts that need not qualify as ideas by satisfying the strict requirement of internally representing a thing.

The repeated use of proprie connects the restriction of the universal statement directly to the treatment of common notions. Descartes first specifies that the statement concerns ideas taken proprie et stricte, then explains that ideas of common notions non sunt ideæ rerum proprie. My renderings, “properly and strictly” and “not properly ideas of things,” preserve the verbal connection. The qualification that identifies the ideas governed by the statement is the same qualification under which common notions are denied the status of ideas of things. Cottingham’s “strict and narrow” and “strictly speaking” make this repetition less visible. Retaining it enables the reader to follow Descartes’s answer without overlooking either part. The claim about ideas being as if of things concerns the proper and strict use, while Descartes’s reply nevertheless expressly recognizes common notions as ideas under the broader use. Common notions are thus acknowledged Cartesian ideas, not cases that an interpreter may discard merely because they are not properly ideas of things.

The words autem hîc, which I render as “however” and “here,” identify how Descartes answers Burman’s proposed counterexample. Burman introduces the idea of nothing against the requirement that ideas be as if of things. Descartes’s reply calls that idea negative and scarcely an idea, then adds, in my translation, “the author, however, takes ‘idea’ here properly and strictly” (AT V 153; CSMK III 338). “However” marks the contrast between the idea Burman introduces and the strict ideas covered by the requirement. “Here” identifies the particular statement being defended, rather than declaring that Descartes always uses “idea” strictly. The reply therefore answers Burman by restricting the requirement’s application. It does not preserve an unrestricted requirement by turning Nothing into a thing represented by the idea. This matters because treating Nothing as a represented object would assume that the idea of nothing must contain a represented thing, although Burman introduces it precisely as an idea that is not of a thing. Descartes’s reply answers the objection by identifying the statement’s restricted scope. The requirement that ideas be as if of things applies to ideas understood properly and strictly. An idea of nothing, described in the reply as negative and scarcely an idea, therefore does not refute that requirement merely because it can be called an idea in a broader sense. If Descartes meant that everything he calls an idea must be of a thing, the idea of nothing would be a counterexample. But his reply limits that requirement to strict ideas, so a thought called an idea only in the broader sense does not contradict it.

Finally, “but then ‘idea’ is taken more broadly” preserves the comparison expressed by latius (“more broadly”) by relating the use that includes common notions to the proper and strict use just identified. The word “idea” applies to common notions even though Descartes’s reply expressly denies that they are properly ideas of things. Understanding a principle such as nothing comes from nothing is a conscious thought, and its recognition as an idea does not require treating the principle itself as a represented thing comparable to the lion represented in an idea of a lion. Cottingham’s “a rather extended use” can suggest an unusual stretching of terminology because “extended” invites the reader to regard the application to common notions as a departure from the term’s usual application, while “rather” can reinforce that impression. My “taken more broadly” presents their inclusion directly as a recognized broader use of “idea.” Common notions consequently remain among the mental states that must be examined when assessing the claim that every Cartesian idea internally represents something. Excluding them from strict ideas does not exempt a defender of that universal claim from explaining what they allegedly represent.

Read in the proper and strict sense specified by the reply, the statement retains its universal force. Every strict idea must be as if of a thing. What the reply rejects is the assumption that this requirement extends to every thought admitted under Descartes’s broader use of “idea.” Common notions provide the express contrast, since he acknowledges them as ideas while denying that they are properly ideas of things.

With the idea of nothing, Descartes is more guarded. He describes it as only negative and scarcely an idea, so it should not be presented as an unqualified example of the broader use on exactly the same footing as common notions. Nevertheless, he does not answer Burman by supplying an entity called Nothing as its represented object. Instead, he specifies that the requirement Burman challenges concerns strict ideas. The objection therefore gives him a reason to clarify which ideas the statement includes, rather than to insist that the idea of nothing represents a thing after all.

Common notions receive a more explicit acknowledgment as ideas. Their inclusion shows why “idea” and “idea of a thing” cannot be substituted for one another throughout Descartes’s writings. An interpreter who makes that substitution would impose the strict requirement on thoughts that Descartes’s reply expressly places under a broader use.

Evidence for this distinction also appears in the published Principles of Philosophy. Articles I.48–49 distinguish things and their affections from eternal truths or common notions. In I.49, Descartes gives the principle that nothing comes from nothing as an example and explains that it is not considered an existing thing or a mode of a thing. (AT VIIIA 23–24; CSM I 208–209) Understanding this principle therefore need not consist in representing the truth itself as a thing comparable to a lion or a triangle.

Consider, however, what is involved in understanding a truth about a triangle. Even if the connection understood adds no further represented thing, the thought may still contain objectively real content representing the triangle. That complete thought would retain objective reality. To establish a counterexample to universal objective reality, I must therefore examine the concepts through which a common notion is understood as well as the principle considered as a whole. In Chapter Three, I argue that the idea of nothing and the common notion ex nihilo nihil fit (“From nothing, nothing comes”) lack objective reality, including in the concepts involved in understanding them.

We can thus reject the unrestricted reading of AT VII 44 without assuming that the reply to Burman has already established the absence of objective reality in every common notion. His reply specifies which sense of “idea” governs the requirement. Determining whether a particular common notion contains any objective reality requires examining what is actually thought when that principle is understood.

For the argument of Section 2.5, the exchange supplies corroborating evidence. Burman challenges the requirement at AT VII 44, rather than the earlier claim about objective being at AT VII 42. Descartes’s reported answer shows that an apparently universal statement about ideas can concern strict ideas exclusively. My argument for reading AT VII 42 that way rests on the comparison of represented contents and their causes. The exchange supports the distinction used in that argument without substituting for the examination of the earlier passage.

Heat and cold still demand an explanation because the apparently universal statement occurs in Descartes’s discussion of those sensations. He calls them ideas while considering how they can present a non-thing as if a thing. Answering Burman’s objection about nothing and common notions does not by itself identify what a cold sensation contains, and calling that sensation an idea does not establish that it is a strict idea with objective reality. To explain the materially false sensory case, I must also consider Descartes’s account in the Fourth Replies.

On my account, felt coolness is a positive phenomenal determination whose experienced character can incline the perceiver toward a false judgment before assent. The explanation must show how a sensation can present coolness as bodily located and encourage its attribution to a body without already containing an objectively real representation of that bodily quality. Neither the wording at AT VII 44 nor the reply to Burman identifies such content within the sensation. A defender of universal internal representation must therefore explain what the sensation represents and how that representation belongs to the experience. The next section examines the proposal that every broad-sense idea contains a strict representational aspect.

2.7. Does Every Broad Idea Have a Strict Representational Aspect?

Steven Nadler’s interpretation accepts Descartes’s broad use and attributes strict representational content to every idea included under it. His claim is explicit:

At the core of every idea in the broad sense is an idea in the strict sense, giving it a specific content or referent. ¹²Steven Nadler, “The Doctrine of Ideas,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 87. The surrounding discussion distinguishes broad and strict uses and illustrates the proposed strict core with desire for an ice cream cone.

Desire for an ice cream cone provides Nadler’s example of a broad-sense idea with a strict representational core. Like fear of a lion, it has a represented object together with a volitional or affective character. We agree that such thoughts contain represented content, and we also agree that Descartes calls sensations ideas. Citing cold as an idea therefore establishes a classification Nadler already accepts. To challenge his interpretation, I must explain why that sensation lacks the represented content he attributes to every broad-sense idea.

Even if “core” describes an aspect of one mental mode, rather than a second mental entity, Nadler’s universal claim still requires an answer for each disputed case. What subject matter is represented in a cold sensation, in the idea of nothing, or in the common notion that nothing comes from nothing? Identifying the ice cream cone desired or the lion feared answers that question for those thoughts. It leaves the proposed counterexamples to be examined on their own terms.

Later in his account, Nadler links the Third Meditation’s strict classification to its statement that objective being belongs to ideas by nature. On this reading, objective reality defines ideas and gives them represented content.¹³Nadler, “The Doctrine of Ideas,” 91–92, especially 92. The defining-feature passage is on printed page 92 in the volume. It juxtaposes the strict classification at AT VII 37 with the claim about objective being by nature at AT VII 42. I accept that connection for strict ideas. Extending it to every broad-sense idea depends on his further claim that each contains a strict representational core. Reading the two Cartesian passages together does not independently establish that claim.

Comparable universal claims appear in Vere Chappell and Lawrence Nolan. Chappell pairs every materially considered idea with an objectively considered idea; Nolan attributes formal and objective reality to every idea.¹⁴Vere Chappell, “The Theory of Ideas,” in Essays on Descartes’ Meditations, edited by Amélie Oksenberg Rorty (Berkeley: University of California Press, 1986), 177–98, especially 178–79; Lawrence Nolan, “The Ontological Status of Cartesian Natures,” Pacific Philosophical Quarterly 78, no. 2 (1997): 169–94, especially 174–76. See Chapter One, Table 1.1, VC1–3 and LN1–5. Both extend the distinction between mental occurrence and represented content to every idea. Applied to a cold sensation, their claims require it to have represented content as well as an actual occurrence in the mind; recognizing those two features in an idea of a lion does not establish their presence in the sensation.

For a strict idea, I accept both considerations. An idea of a lion is an actual modification of mind whose content represents the lion, and distinguishing these features does not make them independently existing things. What remains to be justified is their attribution to every broad-sense idea, including sensations whose content is in dispute.

Consider what someone might call an “aspect” of feeling cold in the left hand. The felt coolness and its experienced location characterize the sensation itself, whereas being caused by contact with ice describes its relation to a body. Having a bodily cause does not prevent the coolness from being an internal qualitative determination of the experience. But identifying that cause does not also establish an internal representation of the ice or its corporeal structure. Nor does acknowledging the sensation’s felt location settle whether the hand is internally represented. Each proposed representation requires an explanation of how its subject matter is present within the sensation; the word “aspect” supplies no such explanation on its own.

A triangle need not have a physical instance to have objective being in an idea, as Nolan’s account of mathematical natures explains. With phenomenal coolness, however, the disputed question is whether there is any internally represented object to begin with. A represented geometrical nature remains represented even without an external instance. That account does not establish a comparable object within a cold sensation simply because the sensation, too, is a thought.

In David Cunning’s account, sensations lack objective reality and do not qualify as ideas.¹⁵David Cunning, “Descartes on Sensations and Ideas of Sensations,” in An Anthology of Philosophical Studies (Athens: ATINER, 2006), 17–32, especially 18, 23, and 27. See Chapter One, Table 1.1, DC1–5. By restricting ideas to representational thoughts, Cunning can maintain that every idea represents something while placing sensations outside the class to which that universal claim applies. But Descartes also uses “idea” broadly enough to include sensations. Excluding them by definition therefore does not establish that the universal claims hold for Cartesian ideas understood in this broader sense. I agree with Cunning that secondary-quality sensations lack objective reality, but reject his exclusion of those sensations from the category of Cartesian ideas.

Taken together, several Cartesian texts give the universalist a plausible argument. Descartes identifies strict ideas through represented subject matter, describes willing and fearing as including an apprehended object, and explains the causes of represented contents. His broad definition also includes all immediately experienced thoughts. The proposed conclusion is that every thought covered by that definition contains a strict representational aspect. To establish it, the universalist must show why the structure found in fear of a lion and desire for an ice cream cone also belongs to sensations and the other disputed cases.

More directly, Descartes expressly attributes objective reality to ideas produced by the sensory faculty in the Sixth Meditation’s proof concerning corporeal causes. (AT VII 79; CSM II 55) Why, then, deny objective reality to a cold sensation? My answer distinguishes the cognition of a sensible body, which can include objectively real content representing corporeal substance, from its qualitative sensory aspect. On my interpretation, the represented corporeal content belongs to the intellectual aspect of that cognition. Chapter Four defends this reading of the proof. The distinction requires an argument identifying where the represented content occurs. Simply naming two aspects of the cognition would leave the objection unanswered.

A person’s ignorance of the physical cause of a sensation offers no shortcut to proving that it lacks objective reality. John Carriero holds that sensations represent corporeal structure obscurely and confusedly.¹⁶John Carriero, Between Two Worlds: A Reading of Descartes’s Meditations (Princeton, NJ: Princeton University Press, 2009), 20–21, 158, 405, and 409. See Chapter One, Table 1.1, JC1–4. A person’s inability to describe that structure distinctly is compatible with his proposal. I must show why the proposed structure fails to occur as represented content within the sensation. The general definition of an idea and the account of strict ideas do not decide that question in advance.

Smith’s Inconsistent Triad and Sensations as Broad-Sense Ideas

Whether sensations count as ideas determines how Kurt Smith resolves the following inconsistent triad. If “idea” and “representational” keep the same meanings throughout, the three claims cannot all hold:

(1) Ideas are representational,
(2) Sensations are ideas,
(3) Sensations are not representational.¹⁷Kurt Smith, “Descartes’s Ontology of Sensation,” Canadian Journal of Philosophy 35, no. 4 (2005): 563–84, at 563. Smith calls these claims an inconsistent triad and argues that Descartes accepts (1) and (3) but rejects (2).

Smith retains (1) and (3) and rejects (2), concluding that sensations do not qualify as ideas because they do not represent. I reject (1) when it attributes internal representation to every broad-sense Cartesian idea. Secondary-quality sensations qualify as ideas while lacking internal representation. Restricting (1) to strict ideas preserves their representational character and allows sensations to remain ideas under Descartes’s broader use.

Under Descartes’s broad definition, any immediately perceived thought qualifies as an idea. Consistently with that definition, the Fourth Replies calls color and cold ideas while considering their failure to represent anything real. (AT VII 160; CSM II 113; AT VII 234; CSM II 163) The sensation itself qualifies. It need not first become the represented object of a separate reflective thought.

How we read (3) also depends on the kind of representation being denied. I accept it for secondary-quality sensations if it denies internal representation. I reject it if it denies every kind of representation, since these sensations externally represent their causes through lawful natural signification. A cold sensation can signify a bodily property or interaction without containing an objectively real internal representation of what it signifies.

By rejecting (2), Smith removes sensations from the class of ideas to which he attributes representation. Descartes’s broad use, however, includes these experiences whether or not they possess objective reality, and I dispute the grounds for excluding them. This disagreement affects which experiences can count against a universal theory of Cartesian ideas. If sensations qualify as Cartesian ideas but lack objectively real content, they disprove the stronger thesis. If they also lack internal representation, they disprove the weaker thesis. Excluding them by restricting the meaning of “idea” leaves unanswered both Descartes’s broader usage and the arguments about sensory content. The decisive question is what these acknowledged ideas contain. Chapters Four and Five develop the arguments that bare secondary-quality sensations lack objective reality and internal representation.

2.8. What Does Purporting to Represent Add?

A sensation might be said to purport to represent a thing even if it fails to represent anything real. Could this preserve the universal attribution of internal representation? Michael Della Rocca’s formulation makes the possibility explicit:

And, for Descartes, all ideas are representations; they are of things or at least they purport to be of things. ¹⁸Michael Della Rocca, “Judgment and Will,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 142–59, at 146.

In the passage containing this claim, Della Rocca discusses how the intellect presents contents for the will to affirm or deny, or about which to suspend judgment. He immediately cites the Third Meditation’s strict characterization of ideas. I therefore read “representations” here as referring to what comes before the mind for judgment, rather than to a sensation’s lawful signification of its bodily cause.

Which ideas does “all” include? If Della Rocca means strict ideas, the claim does not establish representation in every broad-sense thought. If he includes every idea Descartes recognizes, he must explain the supposed representation in the idea of nothing, common notions, and secondary-quality sensations.

An imagined chair illustrates one use of “purport to be” of a thing. Even if no corresponding chair exists, the thought still has the chair as its represented subject matter. Allowing the representation to lack an external counterpart leaves its content intact. With the idea of nothing and some common notions, the question comes earlier. Must understanding them involve an internally represented thing at all?

To answer by calling Nothing a nonexistent object would already attribute the positive subject matter I dispute. Similarly, understanding that “nothing comes from nothing” does not require a purported object corresponding to the principle as a whole. The reply to Burman and Principles I.48–49 distinguish these thoughts from ideas of things properly so called. As Chapter Three argues, the question of objective reality must be answered by examining what is understood, including the concepts through which the principle is thought.

Why does a thought purport to present a chair rather than a tree? Any account of internal representation must explain what determines that subject matter. Saying that the thought occurs as a conscious mental mode establishes its formal reality, but does not explain why it represents this thing rather than another. A causal connection can account for the thought’s occurrence or its natural signification; an account of internal representation must show how the subject matter is present through the thought’s content.

On my account, objectively real content performs this explanatory work. The chair has objective being in the idea as its representatum, whether or not a corresponding chair exists outside thought. If the only failure is the absence of an external chair, describing the idea as merely purporting to represent does not deprive it of objective reality.

Distinguishing the mental mode, its content, and the chair it represents does not require three independently existing entities. These terms identify the thought that occurs, the content through which it represents, and its representatum, or thing represented. Nor must the mode represent itself in order to represent the chair. The distinction explains why an appeal to formal reality alone is insufficient here. An actually occurring thought still requires determinate content to be an internal representation of a chair.

For my interpretation, the harder challenge comes from materially false sensory ideas said to represent something despite lacking objective reality. Claudia Lorena García defends such an account.¹⁹Claudia Lorena García, “Descartes: Ideas and the Mark of the Mental,” History of Philosophy & Logical Analysis 3, no. 1 (2000): 21–53, especially 38 and 46. See Chapter One, Table 1.1, CLG1 and CLG5–6. Whether García’s account supports the weaker universal thesis depends on whether the representation she attributes to sensations is internal. If the representation García attributes to these sensations is internal, her interpretation challenges my claim that internal representation requires objectively real content. I cannot answer her by defining internal representation in a way that excludes her proposal from the outset. I must independently establish why Descartes’s account of objective reality explains the internal representation of a determinate representatum and why the proposed alternatives fail. For strict ideas, objective reality is the internally contained representational content through which the representatum is present in the intellect. The representatum has objective being in the idea without existing there as the formally real thing itself, and its ontological rank determines the degree of objective reality. This supplies a positive account of determinate internal representation. García’s alternative must explain how a sensation can internally represent something without the objectively real content that performs this role, and my argument must examine that explanation on its own terms.

A cold sensation brings several proposed objects of representation into view. Its experienced coolness is a formally real qualitative determination of the sensory mode, but calling that quality phenomenally present does not explain whether the mode also represents it. If coolness itself is the proposed representatum, the account must show how experiencing it amounts to representing it. Felt bodily location presents a further challenge. Coolness experienced in the left hand differs from coolness experienced in the right, and that difference belongs to the sensation before the person makes a judgment about its location. A defender of internal representation may therefore propose that the sensation represents the hand, or coolness as located there. I must answer that proposal by examining the experience of bodily location, rather than assuming that a formally real mental quality must be felt without a location. Finally, the ice or its corporeal structure might be proposed as the represented subject matter. The sensation’s lawful connection with its bodily cause explains how it signifies that cause, but does not establish that the cause has objective being within the sensation. Experienced coolness, its felt location, and its natural signification thus raise distinct questions about what is alleged to be internally represented. Each requires an account of how the proposed subject matter is present in the experience.

My argument in Chapter Five therefore cannot proceed simply from the absence of objective reality in cold sensations to the absence of internal representation. That would assume the connection García disputes. I distinguish the sensory mode’s formal reality, the phenomenal presentation of coolness and its felt bodily location, and the sensation’s external signification. I then examine whether these features establish internal representation without objectively real content. My argument must explain why feeling cold in the left hand does not require an internal representatum, just as it must explain why signifying a bodily cause does not require representing that cause internally. I defend Descartes’s account of objective reality as the explanation of determinate internal representation and argue that phenomenal presentation and natural signification supply no alternative account of such content. On that basis, I defend the claim that bare secondary-quality sensations lack both objective reality and internal representation, although they remain conscious experiences capable of signifying their bodily causes. Della Rocca’s appeal to ideas that represent or purport to represent does not independently establish a different conclusion. His discussion of strict ideas supports their representational character, but extending that account to every broad-sense idea requires identifying represented content in the disputed cases. Allowing a represented thing to lack actual existence does not supply such content where none has been established. Whether the representatum exists outside thought is a different question from whether the mental mode internally represents anything at all.

2.9. Material Falsity and the Cases Still to Be Examined

Having argued that Descartes’s statement about objective being concerns strict ideas, I must still establish the counterexamples to the universal theses. A complete broad-sense idea without objective reality would refute the stronger thesis; one without internal representation would refute the weaker. Fear of a lion will not do, because the complete thought represents the lion even though its affective character is not another represented item. The cases developed in the following chapters must account for the whole thought and show what it contains.

For a materially false cold sensation, the account must also explain how the experience supplies a basis for false judgment before the perceiver assents. I hold that its phenomenal character can do this without containing an internal representation of a bodily quality. Explaining the later false judgment is therefore insufficient. We must identify what is already misleading in the sensation itself.

When a person holds ice, the coolness felt in the hand can encourage attribution of that same qualitative character to the ice. In MF1, this encourages the perceiver to treat a phenomenal quality as though it existed in a physical body. The sensation already has a felt bodily location, but that location, the ice’s role as its cause, and the attribution of phenomenal coolness to the ice must be distinguished. Both the sensation and its cause are real; the error lies in the attribution the experience encourages. In MF3, dropsical thirst similarly prompts drinking as called for by a bodily need when drinking would harm the patient. I take both sensory ideas to lack objective reality, while their phenomenal character supplies a basis for false judgment before assent.

With an idolater’s false idea of a god, MF2, objectively real content is present. The idea represents a supposed finite being and presents that being as genuinely divine. Calling that being a god does not give the idea the objective reality of an infinite substance. Its represented content and confused attribution of divinity together supply a basis for false judgment. (AT VII 233–234; CSM II 163–164) Material falsity therefore cannot be identified with the absence of objective reality.

Across these three cases, the idea presents what is not real in the relevant respect as if it were, thereby supplying a basis for false judgment. Its own character must account for that tendency. A fictional being entertained as fiction does not become materially false merely because someone could misunderstand it. Chapter Seven develops the account through cold, the false god, and dropsical thirst, then examines its further applications.

To establish the proposed counterexamples, Chapter Three examines the idea of nothing and particular common notions, including the concepts used to understand them, and argues that they lack objective reality. Chapter Four turns to secondary-quality sensations. It examines their proposed objects and argues that none supplies objectively real content within the sensation.

Even after objective reality has been denied to sensations, the weaker thesis remains to be answered. Chapter Five argues that secondary-quality sensations also lack internal representation, including the representation alleged to be required by their felt bodily location. It explains how lawful natural signification nevertheless allows them to represent their causes externally. The chapter also develops objectless dread as a complete conscious mental state lacking both objectively real content and internal representation.

Throughout this chapter, the restriction to strict ideas preserves the necessity Descartes attributes to their objective being. His broader definition admits thoughts through immediate awareness, and the reply to Burman expressly distinguishes the broader use from ideas of things properly so called. Fear of a lion shows how strict representational content and experienced fear can belong to one broad-sense idea. Extending that structure to every conscious thought requires more than an example in which it is present.

My Restriction Thesis holds that strict ideas necessarily contain objectively real representational content, while some broad-sense ideas contain none. The passages examined here establish the grounds for reading Descartes’s claim about the nature of ideas as a claim about strict ideas. Whether a cold sensation, the idea of nothing, or a particular common notion lacks objective reality must be established by examining that thought. An account of what makes an idea of a lion represent a lion cannot settle what, if anything, these other ideas contain as represented.

NOTES

OLD BELOW

Chapter Two: Strict and Broad Ideas and the Scope of Objective Reality

Introduction

In the Third Meditation, Descartes states that the objective mode of being belongs to ideas by their own nature (AT VII 42; CSM II 29).²⁰My translation, “by their own nature,” is philosophically preferable to Cottingham’s “by their very nature” as a translation of Descartes’s Latin: “iste modus essendi objectivus competit ideis ex ipsarum naturâ.” (AT VII 42; CSM II 29)

The crucial expression is ex ipsarum naturâ. The emphatic pronoun ipsarum directs attention to the ideas themselves, whose nature grounds the objective mode of being. The preposition ex identifies that nature as the source or ground of the objective mode of being.

There are three reasons to prefer “own”:

(1) Grammatical precision. “Their own nature” makes the reference to the ideas themselves explicit. “Their very nature” is grammatically defensible, but it emphasizes the nature itself rather than its belonging to the ideas.

(2) Philosophical precision. “Own” directs attention to what belongs to the ideas under discussion in virtue of their nature; I identify those ideas as strict ideas from the surrounding argument. It does not, by itself, establish that every mental mode qualifies as an idea possessing objective reality.

(3) Interpretive significance. The universalist interpretation cannot derive its unrestricted scope merely from ipsarum. It must independently establish which cognitions fall within the extension of ideis in this passage.

There is, however, one qualification. The Latin ipsarum can legitimately be translated as “very” when “very” functions as an intensifier of identity rather than degree. My translation makes the reference to the ideas themselves more explicit, but neither rendering determines whether Descartes means every thought he calls an idea or only the strict ideas whose represented contents he has been comparing. Which ideas? If he means every thought he calls an idea, the statement supports the stronger thesis: sensations, appetites, volitions, and every other broad-sense idea must contain objectively real representational content. Yet the statement occurs within an argument comparing ideas according to the things they represent. I argue that Descartes is speaking exclusively about strict ideas here and that his explanation of their nature does not establish what every conscious thought must contain.

A strict idea internally represents a subject matter through objectively real content. The representatum, or item represented, has objective being in the idea, and its ontological rank determines the content’s degree of objective reality. Descartes’s broad use also includes sensations, willing, fearing, and other immediately experienced thoughts. The weaker thesis requires every one of these thoughts to represent something internally; the stronger thesis requires each to contain objectively real content. The universalist must explain why those requirements apply to every thought included under the broad definition. Calling them ideas supplies neither explanation.

The textual question is whether Descartes’s statements attributing objective reality to ideas apply to every Cartesian idea or only to ideas understood properly and strictly as internally representing things. Descartes’s classification of thoughts in the Third Meditation and his subsequent comparison of what different ideas represent provide grounds for restricting the scope of those statements. The reply reported in the Conversation with Burman distinguishes ideas taken “properly and strictly” from the broader use for common notions, providing further evidence that apparently unrestricted statements about ideas of things apply exclusively to strict ideas. (AT V 153; CSMK III 338) Refuting the universal claims also requires examples of complete ideas that lack the content attributed to them. Chapters Four and Five supply those arguments by examining sensations that lack objective reality and by establishing that bare secondary-quality sensations lack internal representation.

2.1. What Qualifies as a Cartesian Idea?

In the geometrical exposition appended to the Second Replies, Descartes defines “idea” as follows.

I understand this term [‘idea’] to mean the form of any given thought, immediate perception of which makes me aware of the thought. (AT VII 160; CSM II 113)²¹Second Replies, Definition II. The definition continues across AT VII 160–61. English quotations follow CSM unless another translation is identified.

Immediate awareness makes the thought qualify as an idea under this definition. Descartes does not add that something must be objectively represented within it. A commentator who attributes represented content to every thought covered by “any given thought” must explain why sensing, willing, fearing, and the other conscious activities all require it.

Definition II continues by distinguishing the form of a thought from a corporeal image depicted in the brain. Descartes does not count the brain image itself as the mental idea. Imaginative thoughts qualify because the mind is immediately aware of them, and the definition extends to other thoughts on the same basis. (AT VII 160–61; CSM II 113) Excluding a bodily picture from the definition leaves open what the conscious thought contains.

Descartes makes the breadth of his usage especially plain when replying to Hobbes. In the Third Replies, he writes that he takes “idea” to refer to whatever the mind immediately perceives, and then gives the reason for including wanting and fearing.

For example, when I want something, or am afraid of something, I simultaneously perceive that I want, or am afraid; and this is why I count volition and fear among my ideas. (AT VII 181; CSM II 127)

Descartes’s reason for counting wanting and fearing as ideas is that they are immediately perceived. The broad use includes the volitional or affective thought itself. It does not reserve “idea” for the representation of the object while excluding the wanting or fearing.

Nor must awareness of fearing be a second act directed at the first. In the Letter to Mersenne of 28 January 1641 (AT III 295; CSMK III 172), Descartes denies that the idea by which willing is known must differ from the willing itself. The person is aware of willing in the act of willing.²² A later reflective thought takes that act of willing as its represented subject matter. Such reflection may contain objectively real content representing the earlier act. But this does not show that the original act of willing represented itself or that its volitional aspect consisted in objectively real representational content.

These examples involve wanting something and fearing something, so they do not show that the complete thoughts lack objective reality. Fearing a lion includes objectively real content representing the lion. What the passage establishes is that the experienced fear also belongs to the thought Descartes calls an idea.

Descartes includes sensations too. In the Fourth Replies (AT VII 234; CSM II 163), he calls color and cold ideas arriving from the senses while considering their failure to represent anything real. Their disputed objective reality gives no reason to exclude them from his broad use of “idea.”²³David Cunning excludes sensations from the category of ideas: “A Cartesian sensation is a mode of mind but not an idea. If it is representational, it is not representational in virtue of having objective reality but in virtue of something else.” (“Descartes on Sensations and Ideas of Sensations,” in An Anthology of Philosophical Studies, vol. I, edited by Patricia Hanna, Adrianne Leigh McEvoy, and Penelope Voutsina, Athens, Greece: Atiner Publishing, 2006, 18)

Cunning later explains his restriction: “Another is that ideas are representational. Indeed, Descartes says that although something can be dubbed a ‘thought’ in a loose sense so long as it is a mode of mind, the term ‘idea’ is to be reserved for modes of mind that are of objects.” (23)

This restriction fits the strict use of “idea” in the Third Meditation (AT VII 37; CSM II 25–26), but Descartes also recognizes a broader use. In the Fourth Replies, he expressly calls colour and cold sensory ideas while considering their failure to represent anything real (AT VII 234; CSM II 163).

I agree with Cunning that secondary-quality sensations lack objective reality, but reject his exclusion of sensations from Cartesian ideas.

Cunning also locates material falsity in an idea of a sensation (18), whereas I argue that a materially false sensation itself has a misleading presentational character before any false judgment is made. His exclusion of sensations therefore differs from my rejection of the universal theses, which concern every idea in Descartes’s broad sense. The dispute is over what these acknowledged sensory ideas contain.

Calling a thought an operation of the intellect does not yet establish that it contains objectively real representational content, because Descartes uses “intellect” more broadly than “pure understanding.” In Principles of Philosophy I.9 (AT VIIIA 7; CSM I 195), he includes understanding, willing, imagining, and sensory awareness among the activities of thought. In I.32, he assigns perception to the intellect and volition to the will, then states that “Sensory perception, imagination and pure understanding are simply various modes of perception.” (AT VIIIA 17; CSM I 204) Sensation therefore belongs to the intellect when “intellect” means the faculty of perception, although it is not an act of pure understanding. Its classification as an intellectual perception identifies the faculty to which it belongs. Whether it contains objectively real content requires a further explanation of what, if anything, it internally represents.

Consider how this distinction applies to fearing a lion and experiencing cold. In fearing a lion, the fear is immediately experienced, while the lion is internally represented through objectively real content. Both belong to one thought, but the fear is a formally real affective determination of the mental mode rather than another internally represented item alongside the lion. A cold sensation likewise has an immediately experienced character: its phenomenal coolness is a formally real qualitative determination of the sensory mode. On my interpretation, however, the bare sensation contains no internal representatum. I defend the absence of objectively real content in Chapter Four and the absence of internal representation in Chapter Five. The coolness is phenomenally presented rather than objectively represented. Descartes’s broad definition of “idea” admits both the fear of a lion and the cold sensation without requiring them to possess the same kind of content. What is immediately experienced in a thought must therefore be distinguished from what, if anything, that thought internally represents.

2.2. Strict Ideas and the Image Analogy

In the Third Meditation, Descartes distinguishes different kinds of thought to determine which can properly be true or false.

Some of my thoughts are as it were the images of things, and it is only in these cases that the term ‘idea’ is strictly appropriate—for example, when I think of a man, or a chimera, or the sky, or an angel, or God. (AT VII 37; CSM II 25)

The Latin description is tanquam rerum imagines, which I render as “as if images of things.” The expressions quædam, “some,” and quibus solis proprie convenit ideæ nomen, “to which alone the name ‘idea’ properly belongs,” identify the thoughts receiving the strict name. Descartes then describes willing, fearing, affirming, and denying as having additional forms. I take their volitional, affective, and judgmental character to be distinguishable from the represented subject matter within the same thought.

An angel and God appear in the list. The image analogy therefore cannot require a visual picture. A strict idea specifies what is thought of, its representatum, which can be understood without a visual likeness or an existing external counterpart.

A chimera, with a lion’s body, a serpent for a tail, and a goat’s neck and head projecting from its back, illustrates why a represented thing’s nonexistence does not deprive its idea of objective reality. No such creature exists, yet Descartes counts its idea among strict ideas. The represented creature has objective being in the thought without existing formally, i.e., actually. Consequently, an idea of a nonexistent creature does not refute the universal attribution of objective reality. A counterexample must be an idea that lacks objectively real representational content, whereas the chimera is objectively present in the idea despite having no formal existence outside it.

In Definition III of the Second Replies, Descartes explains objective reality by reference to represented content.

[…] the being of the thing which is represented by an idea, in so far as this exists in the idea. (AT VII 161; CSM II 113)²⁴Second Replies, Definition III. The initial ellipsis omits the introductory words of the definition.

In an idea of a lion, the lion has objective being insofar as it is represented in the idea. I understand objective reality as the internally contained representational content through which the representatum is present in the intellect. Its degree depends on the representatum’s ontological rank as a mode, a finite substance, or an infinite substance. An idea of a lion represents a finite substance, while the idea of God represents an infinite substance and therefore, on Descartes’s account, contains infinite objective reality.

Calling a cold sensation an idea does not establish that it contains a represented thing as an idea of a lion does. Definition III explains objective reality through what exists in an idea as represented. Extending that account to every broad-sense idea requires identifying represented content in each case.

Both uses of “idea” are deliberate. Descartes reserves the strict name for thoughts that are as if images of things (tanquam rerum imagines), gives the broad definition through immediate awareness in the Second Replies, and defends that broader use against Hobbes by including volition and fear.²⁵The Third Meditation identifies thoughts that receive the name “idea” properly and strictly through their represented subject matter. (AT VII 37; CSM II 25) Definition II in the Second Replies explains an idea through immediate awareness of a thought. (AT VII 160–161; CSM II 113–114) The reply to Hobbes expressly includes volition and fear because they are immediately perceived. (AT VII 181; CSM II 127–128) The Fifth Replies also rejects restricting ideas to images in the imagination. (AT VII 366–367; CSM II 253)

The inclusion of volition and fear establishes that the broad use covers the thought with its experienced volitional or affective character. It does not establish that the complete thought lacks represented subject matter. Fearing a lion includes objectively real content representing the lion, although the experienced fear is not itself another represented item. Nadler can therefore accept these passages while maintaining that every broad-sense idea contains strict representational content. The distinction between the two uses must be preserved, but neither the presence nor the absence of objective reality in every thought admitted under the broad use follows from that distinction alone. Any interpretation must account for the ideas Descartes expressly includes under each use.

2.3. Fearing a Lion and Having a Representational Aspect

Willing and fearing make it necessary to distinguish the thought’s represented subject matter from its volitional or affective character. Such thoughts can internally represent something while also containing features that are not themselves represented. Fearing a lion, for example, involves objectively real content representing the lion together with the formally real experience of fear. Must the affective character also count as objectively real representational content merely because it belongs to the same thought? I argue that it need not. The presence of objective reality within a thought does not make every feature of that thought internally representational. Descartes introduces willing and fearing immediately after his discussion of ideas understood properly and strictly:

Other thoughts have various additional forms: thus when I will, or am afraid, or affirm, or deny, there is always a particular thing which I take as the object of my thought, but my thought includes something more than the likeness of that thing. Some thoughts in this category are called volitions or emotions, while others are called judgements. (AT VII 37; CSM II 26)

Consider fearing a lion. The thought contains objectively real content representing a lion, and the fear is an experienced affective determination of that thought. Merely thinking of a lion need not involve fearing it. The fear makes a difference to how the thought occurs without itself becoming another representatum.

The representation of the lion and the experienced fear belong to one mental mode. Descartes’s contrast between the likeness of the thing and what the thought includes beyond that likeness supports distinguishing these features. It gives no reason to turn them into two independently occurring acts.

The word semper, “always,” gives a universalist a substantial textual argument. Descartes assigns an apprehended object to willing, fearing, affirming, and denying as he describes them in this passage. Fear of a lion fits that description. The question is whether these examples establish the same structure in every conscious thought, including a bare cold sensation, objectless dread, and the idea of nothing.

A cold sensation has felt coolness. To extend the account of fearing a lion to this sensation, a commentator must identify what it internally represents and explain why the feeling requires that representation. Chapter Four argues that secondary-quality sensations lack objectively real content. Chapter Five argues that they also lack internal representation and explains how their lawful connections with bodily causes allow them to function as natural signs.

Fearing a lion would not provide a counterexample, even though the fear itself is non-representational. The thought also contains objectively real content representing the lion, and attending only to its affective character does not establish that the complete idea lacks objective reality.

Objectless dread provides the counterexample that fearing a lion does not. Its felt character constitutes a complete affective experience without objectively real content representing a lion, danger, or another subject matter. Descartes need not have discussed it for it to qualify under his broad definition: the person immediately experiences this mental state. In Chapter Five, I defend its absence of objective reality and internal representation. An affective state of this kind disproves both universal theses. Descartes’s description of fear directed at an object cannot establish that every affective experience has such an object.

Nadler nevertheless extends the structure illustrated by fearing a lion to every broad-sense idea: “At the core of every idea in the broad sense is an idea in the strict sense, giving it a specific content or referent.”²⁶Steven Nadler, “The Doctrine of Ideas,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 87. He recognizes both uses of “idea” and attributes strict representational content to every thought admitted under the broader use. My objection is to that universal attribution. The textual support for that attribution depends in part on how Descartes moves from his classification of thoughts to his comparison of represented contents.

2.4. Ideas Considered as Thoughts and as Representing Things

Descartes examines why he has taken some ideas to originate outside himself. A natural impulse and the apparent independence of sensory ideas from his will do not establish their origin or resemblance to external things.²⁷Descartes distinguishes “a spontaneous impulse” to believe that ideas resemble external things from knowledge supplied by the natural light. He recalls that natural impulses have previously misled him. (AT VII 38–39; CSM II 26–27)

He also allows that a faculty within himself, not yet fully understood, could produce ideas independently of his will without assistance from external things. Finally, even an external origin would not guarantee resemblance. His sensory and astronomical ideas of the sun differ, and he rejects the sensory appearance as an accurate likeness of the sun. (AT VII 39; CSM II 27) He then announces a different approach:

[…] there is another way of investigating whether some of the things of which I possess ideas exist outside me. (AT VII 40; CSM II 27)

Descartes first considers ideas “quatenus ideae istae cogitandi quidam modi tantùm sunt, non agnosco ullam inter ipsas inaequalitatem,” translated by Cottingham as “In so far as the ideas are simply modes of thought, there is no recognizable inequality among them.” (AT VII 40; CSM II 27) He then considers them “sed, quatenus una unam rem, alia aliam repraesentat, patet easdem esse ab invicem valde diversas.” [But in so far as different ideas represent different things, it is clear that they differ widely.] (AT VII 40; CSM II 28)

In both clauses, quatenus, “insofar as,” specifies how the ideas are being considered. As modifications of a finite mind, they have the same formal rank. Their differences in objective reality emerge when Descartes considers what they represent. Ideas of substances contain more objective reality than ideas of modes or accidents, and the idea of an infinite substance contains more than ideas of finite substances. These differences in represented reality determine how much reality their adequate causes must possess.

An idea of a lion can be considered in both ways. It exists formally as a mental mode and contains objectively real content representing a finite substance. The lion has objective being as represented without existing in the mind as a physical animal. The mental mode and its represented content are distinguishable within this one idea.

The idea of God makes the difference in rank especially clear. The act of thinking remains a finite modification of mind even though its represented subject matter is an infinite substance. Descartes argues that only a cause with infinite reality can account for that objectively real content. (AT VII 40–42; CSM II 28–29)

Vividness does not determine these degrees. An idea’s intensity or complexity can vary without changing the ontological rank of its representatum.

The two uses of quatenus do not establish that some ideas lack objective reality. A universalist could accept both ways of considering an idea and maintain that both apply to every conscious thought. What the passage establishes is how objective reality enters Descartes’s argument: he compares ideas through the different things they represent. The second clause concerns ideas that represent one thing rather than another; it should not be generalized into the separate thesis that every broad-sense idea necessarily has an objectively real representational aspect.

Descartes next asks why an idea contains the particular objective reality it does. Its cause must possess at least as much reality, formally or eminently, as the idea contains objectively. (AT VII 41; CSM II 28–29) Accounting for the occurrence of a finite mental mode therefore does not exhaust the causal question. The cause must also account for what the idea represents.

The “Preface” to the Meditations distinguishes an idea taken materially, as an operation of the intellect, from an idea taken objectively, as the thing represented by that operation. (AT VII 8; CSM II 7) This distinction differs from the broad and strict uses of “idea.” A strict idea is an operation of the intellect and can be considered both materially and objectively. Describing a cold sensation as an operation of the intellect establishes its status as a mental mode; whether it also contains a represented thing is the disputed claim.

A further terminological difference matters in the Fourth Replies. Descartes writes that “when we think of them as representing something we are taking them not materially but formally.” (AT VII 232; CSM II 163) Taking an idea formally in this passage means considering it as representing something. Its formal reality is its actual existence as a modification of mind. The idea of a lion has formal reality as a mental mode and represents the lion through its objectively real content.

The statement that objective being belongs to ideas by their nature occurs within this explanation of represented content and its causes. I read it as expressing what strict ideas necessarily possess. The surrounding argument supplies the reasons for that restriction.

2.5. Why Objective Being Belongs to Strict Ideas by ‘Their Own Nature’

Why does the objective mode of being belong to ideas by their own nature? Descartes answers while explaining why their causes must possess sufficient reality formally or eminently. I translate ex ipsarum naturâ as “by their own nature,” preserving the reference to the ideas themselves. Which ideas he means must be established from the argument. The sentence reads:

Nam quemadmodum iste modus essendi objectivus competit ideis ex ipsarum naturâ, ita modus essendi formalis competit idearum causis, saltem primis & præcipuis, ex earum naturâ. [My preferred translation: For just as this objective mode of being belongs to ideas by their own nature, so the formal mode of being belongs to the causes of ideas—at least the first and principal causes—by their nature.]

[Cottingham’s translation: For just as the objective mode of being belongs to ideas by their very nature, so the formal mode of being belongs to the causes of ideas—or at least the first and most important ones—by their very nature. (AT VII 42; CSM II 29)]²⁸I prefer “by their own nature” for ex ipsarum naturâ because ipsarum emphasizes the ideas themselves. The feminine plural pronoun refers to the ideas mentioned by ideis; its genitive case identifies whose nature Descartes means. Cottingham’s “by their very nature” is defensible, but “their own” makes this reference more direct. Both translations preserve the necessity of what belongs to the ideas by nature.

The concluding ex earum naturâ, “by their nature,” refers to the causes of ideas. Descartes uses the ordinary earum here, whereas the earlier ipsarum is emphatic. Cottingham renders both phrases with “their very nature.” My translation retains the difference.

I translate saltem primis & præcipuis as “at least the first and principal causes.” Primis identifies the first causes; præcipuis identifies the principal or chief causes. Descartes immediately explains why the causal account must reach them: an indefinite succession of ideas originating from other ideas would leave unexplained the source of the reality they contain objectively. The ultimate cause must possess sufficient reality formally or eminently. “Principal” makes this explanatory role clearer than Cottingham’s “most important,” although his wording is lexically defensible. Repeating “causes” also keeps that role explicit.

The idea of God illustrates the issue. Its occurrence is a finite mental mode, but its content represents an infinite substance. The cause must account for that infinite objective reality. The same requirement applies according to the appropriate rank in other cases: an idea of a lion contains the objective reality of a finite substance.

Finally, iste, “this,” points back to the objective mode of being just explained. Cottingham’s “the” leaves that reference implicit. Translating it as “this” makes the continuity of the argument more apparent, but the demonstrative does not by itself restrict the statement to strict ideas. Nor does ipsarum supply that restriction merely by emphasizing the ideas. My identification of them as strict ideas rests on Descartes’s comparison of represented contents and his explanation of the reality their causes must possess. (AT VII 42; CSM II 29)

A strict idea cannot represent a lion without the lion having objective being in it. On my interpretation, its objectively real content accounts for that representation. Descartes’s appeal to the ideas’ nature expresses the necessity of this connection.

He writes ideis, “to ideas,” without adding “strict” or “broad.” The preceding argument identifies the ideas at issue. Descartes has compared them according to what they represent and asked why their different degrees of objective reality require adequate causes. His statement explains how the represented reality exists in those ideas and why its ultimate source cannot be merely another representation.

The comparison begins with thoughts to which the strict name belongs: ideas of a man, a chimera, an angel, and God. (AT VII 37; CSM II 25) Each has represented subject matter, even when no corresponding thing exists outside thought. Descartes can therefore compare their objective contents before establishing the existence of anything beyond the mind. The chimera illustrates why the comparison remains possible: although no such creature exists, its idea still contains represented subject matter.

When Descartes distinguishes ideas as modes of thought from ideas considered according to what they represent, the comparison becomes one of ontological rank. Ideas of finite substances contain more objective reality than ideas of modes, and the idea of an infinite substance contains more than ideas of finite substances. (AT VII 40; CSM II 27–28) Immediate awareness alone gives none of these differences. They arise from what the ideas represent and determine how much reality an adequate cause must possess.

The causal question follows from those differences. Why does an idea contain this particular objective reality? Descartes requires its cause to possess at least as much reality formally or eminently as the idea contains objectively, explaining that objective being is not nothing and cannot come from nothing. (AT VII 41; CSM II 28–29) The reality for which he is seeking a cause is represented content. This does not turn every real feature of a thought into an objectively represented item. In fearing a lion, the fear is formally real as an affective determination of the mental mode, but its reality does not make it another represented object.

Descartes then refers to iste modus essendi objectivus, “this objective mode of being” (AT VII 42; CSM II 29), returning to the way a represented thing exists in the intellect through an idea. The comparison quemadmodum … ita, “just as … so,” connects that objective being with the formal being belonging to the causes of ideas. The represented lion has objective being in the idea, and whatever ultimately accounts for that content must possess sufficient reality formally or eminently. The sentence continues the explanation of represented reality and its cause.

The demonstrative iste does not settle the question of scope. A universalist can accept its backward reference and still claim that every broad-sense idea contains represented content. The reason for restricting the sentence comes from the connected argument: Descartes identifies thoughts that represent things, compares their contents, and explains what can cause them. He has supplied no argument at this point that every sensation or appetite contains such a representation.

Immediately afterward, Descartes again calls ideas images, using veluti quasdam imagines, “like certain images” (AT VII 42; CSM II 29). The reality they contain cannot exceed that of their ultimate sources. An idea of a stone, for example, contains objectively real content representing a corporeal substance, and its cause must possess sufficient reality to account for that content. The renewed image analogy continues the discussion of ideas that represent things. It gives no indication that Descartes has shifted to every thought included under his broad definition.

The comparison of represented contents leads into the statement about the ideas’ nature, and the renewed description of ideas as images follows it. Between these passages, Descartes explains the causes required for the reality contained objectively in ideas. I conclude that his statement characterizes strict ideas. The argument surrounding the sentence supports this reading independently of the counterexamples developed later in the book.

A universalist may accept this account of the passage and add that every broad-sense idea contains a strict representational aspect. Fear of a lion shows how that can happen. Descartes can consider the represented lion when comparing objective reality without counting the fear as another representation. But what plays the role of the lion in a cold sensation or in the idea of nothing? The explanation of fear supplies no answer for those cases.

A defender of the weaker thesis may hold that every broad-sense idea internally represents something while allowing that some lack objective reality. This offers a way to preserve universal internal representation after rejecting universal objective reality. A passage explaining the objective being of represented contents cannot itself establish that representation survives without such content. In Chapter Five, I argue against that separation by examining what secondary-quality sensations are alleged to represent and how the proposed representation could occur.

Restricting the statement to strict ideas preserves the necessity Descartes attributes to their objective being. Every strict idea contains objectively real representational content, while a thought lacking that content can still qualify as an idea under the broad definition through immediate awareness. Descartes makes this broader use explicit in the Third Replies, where he explains to Hobbes that volition and fear count among his ideas because he immediately perceives himself willing or fearing (AT VII 181; CSM II 127). These examples do not establish that the complete thoughts lack objective reality, since fearing a lion also includes represented content concerning the lion. They establish why the experienced willing or fearing belongs within his broad use of “idea.” A thought’s failure to qualify as a strict idea therefore leaves its status as a broad-sense idea intact, provided it meets Descartes’s requirement of immediate awareness.

The causal argument for God’s existence requires the idea of an infinite substance to contain more objective reality than a finite mind can supply formally or eminently. Whether a cold sensation has objective reality does not alter that content or its causal requirement. Descartes’s proof can proceed without attributing objective reality to every sensation, appetite, or act of willing.

Descartes’s causal proof concerns the idea of God and the cause of its infinite objective reality. The proof does not also require a sensation of cold to contain objectively real content. But this leaves open what Descartes believed about the sensation. He could have thought that every idea contains objective reality even though his proof did not require that universal claim. My reason for restricting his statement at AT VII 42 is therefore the argument surrounding it, where he compares ideas according to what they represent and asks what could cause their represented content. To establish that a cold sensation lacks objective reality, I must separately examine what the sensation contains.

My Restriction Thesis holds that strict ideas necessarily contain objectively real representational content, while some broad-sense ideas contain none. An idea of a lion illustrates the first part. It represents the lion through objectively real content, and without that content it would not be a strict idea of a lion. A bare sensation of cold illustrates the second part on the interpretation I defend. Its felt coolness makes it a conscious experience, but experiencing coolness does not itself establish that the sensation internally represents anything. In Chapter Four, I examine what the sensation is alleged to represent and argue that it contains no objectively real content. Chapter Five then addresses the further claim that it might internally represent something even without objective reality. There I argue that bare secondary-quality sensations lack even internal representation and explain how they nevertheless signify their bodily causes. The reading defended in this section establishes why Descartes’s statement about the nature of strict ideas does not settle those questions in the universalist’s favor.

A defender of universal internal representation can nevertheless appeal to Descartes’s statement in the discussion of heat and cold, “And because there can be no ideas except as if of things” (AT VII 44; CSM II 30, my translation). The wording “no ideas except” excludes exceptions, and its occurrence in a discussion of sensations gives the universalist a direct textual basis for claiming that even a cold sensation must represent something. The reply reported in the Conversation with Burman answers the question of scope by specifying that “idea” in this statement is understood properly and strictly, while common notions belong to a broader use (AT V 153; CSMK III 338). The statement therefore governs strict ideas. Extending its requirement to every conscious thought would substitute the broad sense of “idea” for the strict sense expressly identified in the reply. Calling cold a sensory idea does not justify that substitution or establish what the sensation internally represents. Section 2.6 develops this answer to the universalist reading and examines how the restriction bears on Descartes’s discussion of sensations of heat and cold.

2.6. Apparently Universal Language and the Reply to Burman

While considering the obscure sensory ideas of heat and cold, Descartes makes a statement that a defender of universal internal representation can invoke against my interpretation. Its wording excludes exceptions, so the question is which ideas it includes:

Et quia nullæ ideæ nisi tanquam rerum esse possunt. [Cottingham’s translation: “And since there can be no ideas which are not as it were of things. . .”] [My translation: And because there can be no ideas except as if of things.] (AT VII 44; CSM II 30)²⁹My translation of Et quia nullæ ideæ nisi tanquam rerum esse possunt is “And because there can be no ideas except as if of things.” (AT VII 44; CSM II 30) I prefer “as if” to Cottingham’s “as it were” because the choice affects how the reader understands an idea’s being of something. “As it were” readily invites a qualified attribution, whereas “as if” foregrounds the difference between how something appears or behaves and whether it possesses the attributed character. Consider two descriptions of someone’s position in a department. Saying that he is, as it were, the department’s father credits him with a figurative paternal role. Saying that he behaves as if he were the department’s father describes his conduct without granting him that role. He might simply be assuming an authority he lacks. The distinction is between attributing a character in a qualified sense and describing something as if it possessed that character without making the attribution.

Descartes’s discussion of bodily pain supplies a directly relevant example. In the Sixth Meditation, he writes “sentietur dolor tanquam in pede,” which Cottingham translates as “then pain will be felt as if it were in the foot.” (AT VII 88; CSM II 61) Cottingham here renders tanquam as “as if.” The surrounding explanation concerns a bodily disturbance occurring elsewhere along the nerves, or in the brain itself, that produces the same brain motion ordinarily caused by injury to the foot. The resulting pain is nevertheless felt in the foot. Its felt location belongs to the sensory experience itself. The person need not first experience an unlocated pain and then judge that it is in the foot. “As if” expresses this bodily location as experienced without affirming that the bodily disturbance occurs there.

A cold sensation requires the same attention to what is already present in the experience. On my interpretation, phenomenal coolness is a formally real qualitative determination of the sensory mental mode, and it can be felt in the left hand rather than the right. Its being a mental determination does not prevent it from being experienced as bodily located. The sensation can therefore be of cold through its felt coolness and can present that coolness as located in the hand before any judgment is made. But coolness felt in the hand, the ice cube’s causing that experience, and the attribution of the same phenomenal coolness to the ice are different matters. Neither the felt bodily location nor the identification of the ice as the cause establishes that the ice possesses the qualitative character experienced. Nor should the sensory attribution of location simply be identified with an objectively real representation of the hand. Chapter Five examines the argument that felt bodily location requires such an internal representation.

These examples explain the philosophical importance of translating tanquam rerum as “as if of things.” The qualification permits the experience’s own apparent bodily attribution to be acknowledged without treating that attribution as an established account of what the body possesses or what the sensation internally represents. By contrast, “as it were of things” can encourage a reader to assume that a thing is already represented and to qualify only the manner of its representation. For example, the sensation may be taken to contain an obscure representation of coolness in the ice, although the presence of that represented bodily quality is precisely what requires argument. My translation makes this step visible. Just as pain felt as if in the foot does not establish that its bodily cause is in the foot, coolness experienced as if belonging to a body does not establish that the body possesses phenomenal coolness. The force of tanquam is to keep the sensory appearance explicit, so that what is felt is not silently converted into an assertion about the body or an assumption about the sensation’s objectively real content.

Burman challenges this requirement with the idea of nothing. In Cottingham’s translation of the reported exchange, his objection reads, “But we have an idea of nothing, and this is not an idea of a thing.” (AT V 153; CSMK III 338) If the requirement covers everything Descartes calls an idea, this example contradicts it. Descartes’s reply identifies the sense of “idea” in which the requirement holds:

Illa idea est solùm negativa, et vix vocari potest idea; auctor autem hîc sumit ideam proprie et stricte. Aliæ etiam dantur ideæ notionum communium, quæ non sunt ideæ rerum proprie; sed tum idea latius sumitur. [Cottingham’s translation: {Descartes} That idea is purely negative, and can hardly be called an idea. In this passage the author is taking the word ‘idea’ in its strict and narrow sense. We do also have ideas of common notions, which are not, strictly speaking, ideas of things. But this is a rather extended use of the word ‘idea’.]

[My translation: That idea is only negative and can scarcely be called an idea; the author, however, takes “idea” here properly and strictly. There are also ideas of common notions, which are not properly ideas of things; but then “idea” is taken more broadly.] (AT V 153; CSMK III 338)³⁰My translation of the reported reply in the Conversation with Burman, 16 April 1648. This is a report of the exchange, rather than a passage from a work published by Descartes. The AT apparatus identifies Burman’s objection as directed at the statement that ideas can only be as if of things (AT VII 44; CSM II 30). Burman is not directly challenging the earlier statement that objective being belongs to ideas by their nature (AT VII 42; CSM II 29).

Descartes’s reply therefore expressly restricts the statement about ideas being as if of things to strict ideas. It also supports my reading of the earlier statement by showing that Descartes’s apparently universal language about ideas does not automatically include every thought he calls an idea. The argument for restricting AT VII 42 comes from his comparison of represented contents and their causes, as developed in Section 2.5.

My translation preserves several features of the Latin that explain how Descartes’s reported reply to Burman restricts the claim that ideas can only be as if of things. Burman challenges that claim with the idea of nothing. If every idea must be of a thing, an idea of nothing appears to be a counterexample. Descartes’s reply identifies the relevant use of “idea” as proprie et stricte, “properly and strictly,” and then acknowledges ideas of common notions that non sunt ideæ rerum proprie, “are not properly ideas of things.” He explains their inclusion by adding that idea latius sumitur, “‘idea’ is taken more broadly” (AT V 153; CSMK III 338). By distinguishing these uses of “idea,” Descartes limits the requirement to strict ideas while expressly recognizing common notions as ideas in the broader sense. Someone defending universal internal representation cannot therefore assume that every thought Descartes calls an idea must satisfy the requirement he attributes to ideas properly and strictly understood.

I translate proprie et stricte as “properly and strictly,” preserving both terms. Cottingham’s “in its strict and narrow sense” emphasizes that fewer thoughts fall within this use of “idea,” but “properly” directs attention to what qualifies a thought for that use. An idea of a lion is a strict idea because it internally represents a lion. The question is whether every other thought called an idea must likewise contain an internal representation of a thing. Descartes’s reply expressly distinguishes common notions from ideas of things properly so called. Understanding a principle such as nothing comes from nothing therefore cannot simply be treated as another case of representing a thing in the way that the idea of a lion does. The principle qualifies as an idea under the broader use, while its actual content must be examined to determine whether understanding it involves any represented thing. Preserving “properly” makes the reason for the distinction clearer. The broader category includes thoughts that need not qualify as ideas by satisfying the strict requirement of internally representing a thing.

The repeated use of proprie connects the restriction of the universal statement directly to the treatment of common notions. Descartes first specifies that the statement concerns ideas taken proprie et stricte, then explains that ideas of common notions non sunt ideæ rerum proprie. My renderings, “properly and strictly” and “not properly ideas of things,” preserve the verbal connection. The qualification that identifies the ideas governed by the statement is the same qualification under which common notions are denied the status of ideas of things. Cottingham’s “strict and narrow” and “strictly speaking” make this repetition less visible. Retaining it enables the reader to follow Descartes’s answer without overlooking either part. The claim about ideas being as if of things concerns the proper and strict use, while Descartes’s reply nevertheless expressly recognizes common notions as ideas under the broader use. Common notions are thus acknowledged Cartesian ideas, not cases that an interpreter may discard merely because they are not properly ideas of things.

The words autem hîc, which I render as “however” and “here,” identify how Descartes answers Burman’s proposed counterexample. Burman introduces the idea of nothing against the requirement that ideas be as if of things. Descartes’s reply calls that idea negative and scarcely an idea, then adds, in my translation, “the author, however, takes ‘idea’ here properly and strictly” (AT V 153; CSMK III 338). “However” marks the contrast between the idea Burman introduces and the strict ideas covered by the requirement. “Here” identifies the particular statement being defended, rather than declaring that Descartes always uses “idea” strictly. The reply therefore answers Burman by restricting the requirement’s application. It does not preserve an unrestricted requirement by turning Nothing into a thing represented by the idea. This matters because treating Nothing as a represented object would assume that the idea of nothing must contain a represented thing, although Burman introduces it precisely as an idea that is not of a thing. Descartes’s reply answers the objection by identifying the statement’s restricted scope. The requirement that ideas be as if of things applies to ideas understood properly and strictly. An idea of nothing, described in the reply as negative and scarcely an idea, therefore does not refute that requirement merely because it can be called an idea in a broader sense. If Descartes meant that everything he calls an idea must be of a thing, the idea of nothing would be a counterexample. But his reply limits that requirement to strict ideas, so a thought called an idea only in the broader sense does not contradict it.

Finally, “but then ‘idea’ is taken more broadly” preserves the comparison expressed by latius (“more broadly”) by relating the use that includes common notions to the proper and strict use just identified. The word “idea” applies to common notions even though Descartes’s reply expressly denies that they are properly ideas of things. Understanding a principle such as nothing comes from nothing is a conscious thought, and its recognition as an idea does not require treating the principle itself as a represented thing comparable to the lion represented in an idea of a lion. Cottingham’s “a rather extended use” can suggest an unusual stretching of terminology because “extended” invites the reader to regard the application to common notions as a departure from the term’s usual application, while “rather” can reinforce that impression. My “taken more broadly” presents their inclusion directly as a recognized broader use of “idea.” Common notions consequently remain among the mental states that must be examined when assessing the claim that every Cartesian idea internally represents something. Excluding them from strict ideas does not exempt a defender of that universal claim from explaining what they allegedly represent.

Read in the proper and strict sense specified by the reply, the statement retains its universal force. Every strict idea must be as if of a thing. What the reply rejects is the assumption that this requirement extends to every thought admitted under Descartes’s broader use of “idea.” Common notions provide the express contrast, since he acknowledges them as ideas while denying that they are properly ideas of things.

With the idea of nothing, Descartes is more guarded. He describes it as only negative and scarcely an idea, so it should not be presented as an unqualified example of the broader use on exactly the same footing as common notions. Nevertheless, he does not answer Burman by supplying an entity called Nothing as its represented object. Instead, he specifies that the requirement Burman challenges concerns strict ideas. The objection therefore gives him a reason to clarify which ideas the statement includes, rather than to insist that the idea of nothing represents a thing after all.

Common notions receive a more explicit acknowledgment as ideas. Their inclusion shows why “idea” and “idea of a thing” cannot be substituted for one another throughout Descartes’s writings. An interpreter who makes that substitution would impose the strict requirement on thoughts that Descartes’s reply expressly places under a broader use.

Evidence for this distinction also appears in the published Principles of Philosophy. Articles I.48–49 distinguish things and their affections from eternal truths or common notions. In I.49, Descartes gives the principle that nothing comes from nothing as an example and explains that it is not considered an existing thing or a mode of a thing. (AT VIIIA 23–24; CSM I 208–209) Understanding this principle therefore need not consist in representing the truth itself as a thing comparable to a lion or a triangle.

Consider, however, what is involved in understanding a truth about a triangle. Even if the connection understood adds no further represented thing, the thought may still contain objectively real content representing the triangle. That complete thought would retain objective reality. To establish a counterexample to universal objective reality, I must therefore examine the concepts through which a common notion is understood as well as the principle considered as a whole. In Chapter Three, I argue that the idea of nothing and the common notion ex nihilo nihil fit (“From nothing, nothing comes”) lack objective reality, including in the concepts involved in understanding them.

We can thus reject the unrestricted reading of AT VII 44 without assuming that the reply to Burman has already established the absence of objective reality in every common notion. His reply specifies which sense of “idea” governs the requirement. Determining whether a particular common notion contains any objective reality requires examining what is actually thought when that principle is understood.

For the argument of Section 2.5, the exchange supplies corroborating evidence. Burman challenges the requirement at AT VII 44, rather than the earlier claim about objective being at AT VII 42. Descartes’s reported answer shows that an apparently universal statement about ideas can concern strict ideas exclusively. My argument for reading AT VII 42 that way rests on the comparison of represented contents and their causes. The exchange supports the distinction used in that argument without substituting for the examination of the earlier passage.

Heat and cold still demand an explanation because the apparently universal statement occurs in Descartes’s discussion of those sensations. He calls them ideas while considering how they can present a non-thing as if a thing. Answering Burman’s objection about nothing and common notions does not by itself identify what a cold sensation contains, and calling that sensation an idea does not establish that it is a strict idea with objective reality. To explain the materially false sensory case, I must also consider Descartes’s account in the Fourth Replies.

On my account, felt coolness is a positive phenomenal determination whose experienced character can incline the perceiver toward a false judgment before assent. The explanation must show how a sensation can present coolness as bodily located and encourage its attribution to a body without already containing an objectively real representation of that bodily quality. Neither the wording at AT VII 44 nor the reply to Burman identifies such content within the sensation. A defender of universal internal representation must therefore explain what the sensation represents and how that representation belongs to the experience. The next section examines the proposal that every broad-sense idea contains a strict representational aspect.

2.7. Does Every Broad Idea Have a Strict Representational Aspect?

Steven Nadler’s interpretation accepts Descartes’s broad use and attributes strict representational content to every idea included under it. His claim is explicit:

At the core of every idea in the broad sense is an idea in the strict sense, giving it a specific content or referent. ³¹Steven Nadler, “The Doctrine of Ideas,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 87. The surrounding discussion distinguishes broad and strict uses and illustrates the proposed strict core with desire for an ice cream cone.

Nadler illustrates the claim with a desire for an ice cream cone. Like fear of a lion, the desire has a represented object and a further volitional or affective character. We agree that such thoughts contain represented content. We also agree that Descartes calls sensations ideas. A passage in which Descartes calls cold an idea supports a classification Nadler already accepts. My argument must also answer his claim about what the sensation contains.

The word “core” need not imply a second mental entity. Even if it describes a representational aspect of one mental mode, Nadler’s universal claim remains. What is the represented subject matter in a cold sensation, in the idea of nothing, or in the common notion that nothing comes from nothing? The represented objects of desire and fear answer the question for those thoughts. Each proposed counterexample requires its own answer.

Nadler later joins the Third Meditation’s strict classification to its statement that objective being belongs to ideas by nature. He treats objective reality as a defining feature of ideas and as what gives them represented content.³²Nadler, “The Doctrine of Ideas,” 91–92, especially 92. The defining-feature passage is on printed page 92 in the volume. It juxtaposes the strict classification at AT VII 37 with the claim about objective being by nature at AT VII 42. I accept that connection for strict ideas. Extending it to every broad-sense idea depends on his further claim that each contains a strict representational core. Reading the two Cartesian passages together does not independently establish that claim.

Vere Chappell and Lawrence Nolan also make universal claims. Chappell pairs every materially considered idea with an objectively considered idea; Nolan attributes formal and objective reality to every idea.³³Vere Chappell, “The Theory of Ideas,” in Essays on Descartes’ Meditations, edited by Amélie Oksenberg Rorty (Berkeley: University of California Press, 1986), 177–98, especially 178–79; Lawrence Nolan, “The Ontological Status of Cartesian Natures,” Pacific Philosophical Quarterly 78, no. 2 (1997): 169–94, especially 174–76. See Chapter One, Table 1.1, VC1–3 and LN1–5. Their position is stronger than the observation that an idea’s mental occurrence and represented content can be distinguished.

I accept both considerations for a strict idea. An idea of a lion is an actual modification of mind whose content represents the lion, and distinguishing these features does not make them independently existing things. The disputed step is applying both considerations to every broad-sense idea.

Calling something an “aspect of a sensation” does not establish that it is internal to the sensation. Felt coolness characterizes the experience itself; being caused by contact with ice describes a relation to a body. The coolness remains an internal qualitative determination even though it has a bodily cause. If corporeal structure is also said to be represented within the sensation, an explanation must show how the sensation’s content accounts for that representation. Identifying the structure that produces the feeling does not supply this explanation.

Nolan’s account of mathematical natures shows why an external instance is unnecessary for representation. A triangle can have objective being in an idea even if no corresponding physical triangle exists. But the absence of an external instance is not the difficulty with phenomenal coolness. The question is whether the sensation contains a represented object at all. An account of a represented geometrical nature does not by itself establish a comparable object within the sensation.

A different approach is to deny that sensations possess objective reality while excluding them from the category of ideas altogether. This is David Cunning’s position.³⁴David Cunning, “Descartes on Sensations and Ideas of Sensations,” in An Anthology of Philosophical Studies (Athens: ATINER, 2006), 17–32, especially 18, 23, and 27. See Chapter One, Table 1.1, DC1–5. By restricting ideas to representational thoughts, Cunning can maintain that every idea represents something while placing sensations outside the class to which that universal claim applies. But Descartes also uses “idea” broadly enough to include sensations. Excluding them by definition therefore does not establish that the universal claims hold for Cartesian ideas understood in this broader sense. I agree with Cunning that secondary-quality sensations lack objective reality, but reject his exclusion of those sensations from the category of Cartesian ideas.

The universalist can combine several texts into a plausible argument. Descartes identifies strict ideas through represented subject matter, describes willing and fearing as including an apprehended object, and explains the causes of represented contents. His broad definition includes all immediately experienced thoughts. To conclude that every one of those thoughts contains a strict representational aspect, however, the universalist must show that the structure found in the object-directed examples extends to all the thoughts covered by that definition.

The Sixth Meditation supplies a more direct difficulty. In the proof concerning corporeal causes, Descartes expressly attributes objective reality to ideas produced by the sensory faculty. (AT VII 79; CSM II 55) I distinguish the cognition of a sensible body, which can include objectively real content representing corporeal substance, from its qualitative sensory aspect. On my interpretation, the represented corporeal content belongs to the intellectual aspect of that cognition. Chapter Four defends this reading of the proof. The distinction requires an argument identifying where the represented content occurs. Simply naming two aspects of the cognition would leave the objection unanswered.

Nor does ignorance of a sensation’s physical cause prove the absence of objective reality. John Carriero holds that sensations represent corporeal structure obscurely and confusedly.³⁵John Carriero, Between Two Worlds: A Reading of Descartes’s Meditations (Princeton, NJ: Princeton University Press, 2009), 20–21, 158, 405, and 409. See Chapter One, Table 1.1, JC1–4. A person’s inability to describe that structure distinctly is compatible with his proposal. I must show why the proposed structure fails to occur as represented content within the sensation. The general definition of an idea and the account of strict ideas do not decide that question in advance.

Smith’s Inconsistent Triad and Sensations as Broad-Sense Ideas

The classification of sensations also determines how Kurt Smith resolves the following inconsistent triad. With “idea” and “representational” used in the same senses throughout, the three claims cannot all hold:

(1) Ideas are representational,
(2) Sensations are ideas,
(3) Sensations are not representational.³⁶Kurt Smith, “Descartes’s Ontology of Sensation,” Canadian Journal of Philosophy 35, no. 4 (2005): 563–84, at 563. Smith calls these claims an inconsistent triad and argues that Descartes accepts (1) and (3) but rejects (2).

Smith retains (1) and (3) and rejects (2), concluding that sensations do not qualify as ideas because they do not represent. I reject (1) when it attributes internal representation to every broad-sense Cartesian idea. Secondary-quality sensations qualify as ideas while lacking internal representation. Restricting (1) to strict ideas preserves their representational character and allows sensations to remain ideas under Descartes’s broader use.

Descartes’s broad definition includes any immediately perceived thought, and the Fourth Replies calls color and cold ideas while considering their failure to represent anything real. (AT VII 160; CSM II 113; AT VII 234; CSM II 163) The sensation itself qualifies. It need not first become the represented object of a separate reflective thought.

The sense of “representational” also matters for (3). I reject it if it denies every kind of representation to secondary-quality sensations, since these sensations externally represent their causes through lawful natural signification. A cold sensation can signify a bodily property or interaction without internally containing an objectively real representation of what it signifies. If (3) denies internal representation, I accept it for secondary-quality sensations.

Smith’s rejection of (2) excludes sensations from the class of ideas to which he attributes representation. My objection concerns the grounds for that exclusion, since Descartes’s broad use includes sensations whether or not they possess objective reality. This disagreement affects which experiences can count against a universal theory of Cartesian ideas. If sensations qualify as Cartesian ideas but lack objectively real content, they disprove the stronger thesis. If they also lack internal representation, they disprove the weaker thesis. Excluding them by restricting the meaning of “idea” leaves unanswered both Descartes’s broader usage and the arguments about sensory content. The decisive question is what these acknowledged ideas contain. Chapters Four and Five develop the arguments that bare secondary-quality sensations lack objective reality and internal representation.

2.8. What Does Purporting to Represent Add?

A sensation might be said to purport to represent a thing even if it fails to represent anything real. Could this preserve the universal attribution of internal representation? Michael Della Rocca’s formulation makes the possibility explicit:

And, for Descartes, all ideas are representations; they are of things or at least they purport to be of things. ³⁷Michael Della Rocca, “Judgment and Will,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 142–59, at 146.

Della Rocca discusses contents considered by the intellect on which the will can assent, deny, or suspend judgment, and he immediately cites the Third Meditation’s strict characterization. That context supports reading the representation as internal. He is discussing what comes before the mind for judgment, rather than a sensation’s lawful signification of its bodily cause.

Which ideas does “all” include? If Della Rocca means strict ideas, the claim does not establish representation in every broad-sense thought. If he includes every idea Descartes recognizes, he must explain the supposed representation in the idea of nothing, common notions, and secondary-quality sensations.

Allowing an idea merely to “purport to be” of a thing can accommodate the absence of an external counterpart. An imagined chair is still the chair thought of, even when no such chair exists. This possibility leaves represented subject matter in place. The idea of nothing and some common notions raise a different question: must understanding them involve an internally represented thing at all?

Calling Nothing a nonexistent object would supply the positive subject matter whose attribution I dispute. Nor does understanding that “nothing comes from nothing” require a purported object corresponding to the principle as a whole. The reply to Burman and Principles I.48–49 distinguish these thoughts from ideas of things properly so called. Their particular contents, including the concepts through which the principle is understood, determine whether any objective reality remains.

A proposal about internal representation must also explain what determines the subject matter. Why does this thought purport to present a chair rather than a tree? Its existence as a conscious mental mode does not answer that question. A connection with a cause can explain its occurrence or its natural signification; internal representation requires an account of how the subject matter is present through the thought’s content.

On my account, objectively real content explains how the chair is internally represented, with the representatum having objective being in the idea. The absence of a corresponding chair outside thought leaves that content intact. Describing the idea as merely purporting to represent does not remove its objective reality when this is the only failure alleged.

The mental mode represents a chair, its representational content determines what it represents, and the chair is the representatum, or thing represented. Distinguishing the mode, its content, and its subject matter does not require three independently existing entities. Nor must the mode represent itself in order to represent the chair. Identifying its formal reality establishes that it is an actually occurring thought, but its occurrence alone does not explain its determinate representational content.

A more serious challenge to my interpretation is the claim that materially false sensory ideas can represent something even when they lack objective reality. Claudia Lorena García defends such an interpretation.³⁸Claudia Lorena García, “Descartes: Ideas and the Mark of the Mental,” History of Philosophy & Logical Analysis 3, no. 1 (2000): 21–53, especially 38 and 46. See Chapter One, Table 1.1, CLG1 and CLG5–6. Whether García’s account supports the weaker universal thesis depends on whether the representation she attributes to sensations is internal. What matters for the present disagreement is the kind of representation she attributes to sensations. If she means that they internally represent something despite lacking objective reality, her interpretation challenges the connection I defend between internal representation and objectively real content. I cannot answer her by defining internal representation in a way that excludes her proposal from the outset. I must independently establish why Descartes’s account of objective reality explains the internal representation of a determinate representatum and why the proposed alternatives fail. For strict ideas, objective reality is the internally contained representational content through which the representatum is present in the intellect. The representatum has objective being in the idea without existing there as the formally real thing itself, and its ontological rank determines the degree of objective reality. This supplies a positive account of determinate internal representation. García’s alternative must explain how a sensation can internally represent something without the objectively real content that performs this role, and my argument must examine that explanation on its own terms.

Consider a sensation of cold. The experienced coolness is a formally real qualitative determination of the sensory mental mode. It is phenomenally present, but its presence does not establish that the mode internally represents something. Nor does the sensation’s capacity to signify its bodily cause establish internal representation. That signification depends on the lawful connection between the sensation and its cause; it does not require the cause to have objective being within the sensation. What, then, is the proposed internally represented subject matter? If it is the coolness itself, the account must explain how experiencing that coolness amounts to representing it. A sensory mode can occur with a qualitative character, but attributing representation to it requires showing how that character also serves as represented subject matter. If the proposed subject matter is a bodily quality or the bodily cause of the sensation, the lawful connection explains external signification without establishing either as internally represented. These alternatives require different explanations, but each must identify something beyond the sensation’s mere occurrence that accounts for the internal representation attributed to it.

My argument in Chapter Five therefore cannot proceed simply from the absence of objective reality in cold sensations to the absence of internal representation. That would assume the connection García disputes. I distinguish the sensory mode’s formal reality, the phenomenal presentation of its qualitative character, and its external signification, then examine whether any of these features supplies an alternative account of internal representation. I argue that Descartes’s account of objective reality explains determinate internal representation and that neither phenomenal presentation nor natural signification explains an internal representatum in its absence. On that basis, I defend the claim that bare secondary-quality sensations lack both objective reality and internal representation, although they remain conscious experiences capable of signifying their bodily causes. Della Rocca’s appeal to ideas that represent or purport to represent does not independently establish a different conclusion. His discussion of strict ideas supports their representational character, but extending that account to every broad-sense idea requires identifying represented content in the disputed cases. Allowing a represented thing to lack actual existence does not supply such content where none has been established. Whether the representatum exists outside thought is a different question from whether the mental mode internally represents anything at all.

2.9. Material Falsity and the Cases Still to Be Examined

The interpretation defended above restricts Descartes’s statement about objective being to strict ideas. Refuting universal objective reality also requires a complete broad-sense idea that lacks it. Refuting universal internal representation requires a complete idea with no internal representation. Each case must explain the thought that occurs and what it contains.

Material falsity adds a particular demand. A cold sensation must be able to supply a basis for false judgment through its own phenomenal character, before the perceiver assents. On my interpretation, it does so without containing the false internal representation of a bodily quality that I deny. The explanation cannot postpone the sensation’s misleading character until the judgment has already been made.

Felt coolness in MF1 can incline the perceiver to attribute that same qualitative character to the ice, as though a mental quality could exist in a physical body. The sensation and its cause are real. The error lies in the attribution encouraged by the experience. In MF3, dropsical thirst prompts drinking as called for by a bodily need when drinking would harm the patient. I take both sensory ideas to lack objective reality; their misleading character arises through phenomenal presentation before assent.

The false-god case, MF2, works differently. An idolater’s idea contains objectively real content representing a supposed finite being and presents that being as genuinely divine. Calling that being a god does not give the idea the objective reality of an infinite substance. Its represented content and confused attribution of divinity together supply a basis for false judgment. (AT VII 233–234; CSM II 163–164) Material falsity therefore cannot be identified with the absence of objective reality.

In each case, the idea presents what is not real in the relevant respect as if it were and thereby supplies a basis for false judgment. The idea’s own character must explain that basis. A fictional being entertained as fiction does not become materially false merely because someone could misunderstand it. Chapter Seven develops this account across the three cases and examines its further applications.

The counterexamples require separate arguments. Chapter Three defends the absence of objective reality in the idea of nothing and particular common notions, including in the concepts used to understand them. Chapter Four examines the proposed objects of secondary-quality sensations and argues that none supplies objectively real content within the sensation.

The weaker thesis attributes internal representation to every Cartesian idea. Chapter Five argues that secondary-quality sensations lack internal representation, answering interpretations that retain it despite denying them objective reality. Their lawful natural signification explains how they nevertheless externally represent their causes. The chapter also develops objectless dread as a complete conscious mental state lacking both objectively real content and internal representation.

The textual argument for restriction preserves what Descartes attributes to strict ideas by their nature. His broad definition admits thoughts through immediate awareness; the reply to Burman expressly distinguishes that use from ideas of things properly so called. Fearing a lion shows that a broad-sense idea can contain strict representational content together with experienced fear. It gives no reason to attribute the same structure to every other conscious thought.

My Restriction Thesis holds that strict ideas necessarily contain objectively real representational content, while some broad-sense ideas contain none. The passages examined in this chapter establish the grounds for reading Descartes’s claim about the nature of ideas as a claim about strict ideas. Whether a cold sensation, the idea of nothing, or a common notion lacks objective reality must be established from what that particular thought contains. Descartes’s explanation of strict ideas does not supply the missing content.

NOTES

• 1
  My translation, “by their own nature,” is philosophically preferable to Cottingham’s “by their very nature” as a translation of Descartes’s Latin: “iste modus essendi objectivus competit ideis ex ipsarum naturâ.” (AT VII 42; CSM II 29)
  
  The crucial expression is ex ipsarum naturâ. The emphatic pronoun ipsarum directs attention to the ideas themselves, whose nature grounds the objective mode of being. The preposition ex identifies that nature as the source or ground of the objective mode of being.
  
  There are three reasons to prefer “own”:
  
  (1) Grammatical precision. “Their own nature” makes the reference to the ideas themselves explicit. “Their very nature” is grammatically defensible, but it emphasizes the nature itself rather than its belonging to the ideas.
  
  (2) Philosophical precision. “Own” directs attention to what belongs to the ideas under discussion in virtue of their nature; I identify those ideas as strict ideas from the surrounding argument. It does not, by itself, establish that every mental mode qualifies as an idea possessing objective reality.
  
  (3) Interpretive significance. The universalist interpretation cannot derive its unrestricted scope merely from ipsarum. It must independently establish which cognitions fall within the extension of ideis in this passage.
  
  There is, however, one qualification. The Latin ipsarum can legitimately be translated as “very” when “very” functions as an intensifier of identity rather than degree. My translation makes the reference to the ideas themselves more explicit, but neither rendering determines whether Descartes means every thought he calls an idea or only the strict ideas whose represented contents he has been comparing. ↩︎
• 2
  Second Replies, Definition II. The definition continues across AT VII 160–61. English quotations follow CSM unless another translation is identified. ↩︎
• 3
  A later reflective thought takes that act of willing as its represented subject matter. Such reflection may contain objectively real content representing the earlier act. But this does not show that the original act of willing represented itself or that its volitional aspect consisted in objectively real representational content. ↩︎
• 4
  David Cunning excludes sensations from the category of ideas: “A Cartesian sensation is a mode of mind but not an idea. If it is representational, it is not representational in virtue of having objective reality but in virtue of something else.” (“Descartes on Sensations and Ideas of Sensations,” in An Anthology of Philosophical Studies, vol. I, edited by Patricia Hanna, Adrianne Leigh McEvoy, and Penelope Voutsina, Athens, Greece: Atiner Publishing, 2006, 18)
  
  Cunning later explains his restriction: “Another is that ideas are representational. Indeed, Descartes says that although something can be dubbed a ‘thought’ in a loose sense so long as it is a mode of mind, the term ‘idea’ is to be reserved for modes of mind that are of objects.” (23)
  
  This restriction fits the strict use of “idea” in the Third Meditation (AT VII 37; CSM II 25–26), but Descartes also recognizes a broader use. In the Fourth Replies, he expressly calls colour and cold sensory ideas while considering their failure to represent anything real (AT VII 234; CSM II 163).
  
  I agree with Cunning that secondary-quality sensations lack objective reality, but reject his exclusion of sensations from Cartesian ideas.
  
  Cunning also locates material falsity in an idea of a sensation (18), whereas I argue that a materially false sensation itself has a misleading presentational character before any false judgment is made. His exclusion of sensations therefore differs from my rejection of the universal theses, which concern every idea in Descartes’s broad sense. ↩︎
• 5
  Second Replies, Definition III. The initial ellipsis omits the introductory words of the definition. ↩︎
• 6
  The Third Meditation identifies thoughts that receive the name “idea” properly and strictly through their represented subject matter. (AT VII 37; CSM II 25) Definition II in the Second Replies explains an idea through immediate awareness of a thought. (AT VII 160–161; CSM II 113–114) The reply to Hobbes expressly includes volition and fear because they are immediately perceived. (AT VII 181; CSM II 127–128) The Fifth Replies also rejects restricting ideas to images in the imagination. (AT VII 366–367; CSM II 253)
  
  The inclusion of volition and fear establishes that the broad use covers the thought with its experienced volitional or affective character. It does not establish that the complete thought lacks represented subject matter. Fearing a lion includes objectively real content representing the lion, although the experienced fear is not itself another represented item. Nadler can therefore accept these passages while maintaining that every broad-sense idea contains strict representational content. The distinction between the two uses must be preserved, but neither the presence nor the absence of objective reality in every thought admitted under the broad use follows from that distinction alone. ↩︎
• 7
  Steven Nadler, “The Doctrine of Ideas,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 87. ↩︎
• 8
  Descartes distinguishes “a spontaneous impulse” to believe that ideas resemble external things from knowledge supplied by the natural light. He recalls that natural impulses have previously misled him. (AT VII 38–39; CSM II 26–27)
  
  He also allows that a faculty within himself, not yet fully understood, could produce ideas independently of his will without assistance from external things. Finally, even an external origin would not guarantee resemblance. His sensory and astronomical ideas of the sun differ, and he rejects the sensory appearance as an accurate likeness of the sun. (AT VII 39; CSM II 27) ↩︎
• 9
  I prefer “by their own nature” for ex ipsarum naturâ because ipsarum emphasizes the ideas themselves. The feminine plural pronoun refers to the ideas mentioned by ideis; its genitive case identifies whose nature Descartes means. Cottingham’s “by their very nature” is defensible, but “their own” makes this reference more direct. Both translations preserve the necessity of what belongs to the ideas by nature.
  
  The concluding ex earum naturâ, “by their nature,” refers to the causes of ideas. Descartes uses the ordinary earum here, whereas the earlier ipsarum is emphatic. Cottingham renders both phrases with “their very nature.” My translation retains the difference.
  
  I translate saltem primis & præcipuis as “at least the first and principal causes.” Primis identifies the first causes; præcipuis identifies the principal or chief causes. Descartes immediately explains why the causal account must reach them: an indefinite succession of ideas originating from other ideas would leave unexplained the source of the reality they contain objectively. The ultimate cause must possess sufficient reality formally or eminently. “Principal” makes this explanatory role clearer than Cottingham’s “most important,” although his wording is lexically defensible. Repeating “causes” also keeps that role explicit.
  
  The idea of God illustrates the issue. Its occurrence is a finite mental mode, but its content represents an infinite substance. The cause must account for that infinite objective reality. The same requirement applies according to the appropriate rank in other cases: an idea of a lion contains the objective reality of a finite substance.
  
  Finally, iste, “this,” points back to the objective mode of being just explained. Cottingham’s “the” leaves that reference implicit. Translating it as “this” makes the continuity of the argument more apparent, but the demonstrative does not by itself restrict the statement to strict ideas. Nor does ipsarum supply that restriction merely by emphasizing the ideas. My identification of them as strict ideas rests on Descartes’s comparison of represented contents and his explanation of the reality their causes must possess. (AT VII 42; CSM II 29) ↩︎
• 10
  My translation of Et quia nullæ ideæ nisi tanquam rerum esse possunt is “And because there can be no ideas except as if of things.” (AT VII 44; CSM II 30) I prefer “as if” to Cottingham’s “as it were” because the choice affects how the reader understands an idea’s being of something. “As it were” readily invites a qualified attribution, whereas “as if” foregrounds the difference between how something appears or behaves and whether it possesses the attributed character. Consider two descriptions of someone’s position in a department. Saying that he is, as it were, the department’s father credits him with a figurative paternal role. Saying that he behaves as if he were the department’s father describes his conduct without granting him that role. He might simply be assuming an authority he lacks. The distinction is between attributing a character in a qualified sense and describing something as if it possessed that character without making the attribution.
  
  Descartes’s discussion of bodily pain supplies a directly relevant example. In the Sixth Meditation, he writes “sentietur dolor tanquam in pede,” which Cottingham translates as “then pain will be felt as if it were in the foot.” (AT VII 88; CSM II 61) Cottingham here renders tanquam as “as if.” The surrounding explanation concerns a bodily disturbance occurring elsewhere along the nerves, or in the brain itself, that produces the same brain motion ordinarily caused by injury to the foot. The resulting pain is nevertheless felt in the foot. Its felt location belongs to the sensory experience itself. The person need not first experience an unlocated pain and then judge that it is in the foot. “As if” expresses this bodily location as experienced without affirming that the bodily disturbance occurs there.
  
  A cold sensation requires the same attention to what is already present in the experience. On my interpretation, phenomenal coolness is a formally real qualitative determination of the sensory mental mode, and it can be felt in the left hand rather than the right. Its being a mental determination does not prevent it from being experienced as bodily located. The sensation can therefore be of cold through its felt coolness and can present that coolness as located in the hand before any judgment is made. But coolness felt in the hand, the ice cube’s causing that experience, and the attribution of the same phenomenal coolness to the ice are different matters. Neither the felt bodily location nor the identification of the ice as the cause establishes that the ice possesses the qualitative character experienced. Nor should the sensory attribution of location simply be identified with an objectively real representation of the hand. Chapter Five examines the argument that felt bodily location requires such an internal representation.
  
  These examples explain the philosophical importance of translating tanquam rerum as “as if of things.” The qualification permits the experience’s own apparent bodily attribution to be acknowledged without treating that attribution as an established account of what the body possesses or what the sensation internally represents. By contrast, “as it were of things” can encourage a reader to assume that a thing is already represented and to qualify only the manner of its representation. For example, the sensation may be taken to contain an obscure representation of coolness in the ice, although the presence of that represented bodily quality is precisely what requires argument. My translation makes this step visible. Just as pain felt as if in the foot does not establish that its bodily cause is in the foot, coolness experienced as if belonging to a body does not establish that the body possesses phenomenal coolness. The force of tanquam is to keep the sensory appearance explicit, so that what is felt is not silently converted into an assertion about the body or an assumption about the sensation’s objectively real content. ↩︎
• 11
  My translation of the reported reply in the Conversation with Burman, 16 April 1648. This is a report of the exchange, rather than a passage from a work published by Descartes. The AT apparatus identifies Burman’s objection as directed at the statement that ideas can only be as if of things (AT VII 44; CSM II 30). Burman is not directly challenging the earlier statement that objective being belongs to ideas by their nature (AT VII 42; CSM II 29).
  
  Descartes’s reply therefore expressly restricts the statement about ideas being as if of things to strict ideas. It also supports my reading of the earlier statement by showing that Descartes’s apparently universal language about ideas does not automatically include every thought he calls an idea. The argument for restricting AT VII 42 comes from his comparison of represented contents and their causes, as developed in Section 2.5.
  
  My translation preserves several features of the Latin that explain how Descartes’s reported reply to Burman restricts the claim that ideas can only be as if of things. Burman challenges that claim with the idea of nothing. If every idea must be of a thing, an idea of nothing appears to be a counterexample. Descartes’s reply identifies the relevant use of “idea” as proprie et stricte, “properly and strictly,” and then acknowledges ideas of common notions that non sunt ideæ rerum proprie, “are not properly ideas of things.” He explains their inclusion by adding that idea latius sumitur, “‘idea’ is taken more broadly” (AT V 153; CSMK III 338). By distinguishing these uses of “idea,” Descartes limits the requirement to strict ideas while expressly recognizing common notions as ideas in the broader sense. Someone defending universal internal representation cannot therefore assume that every thought Descartes calls an idea must satisfy the requirement he attributes to ideas properly and strictly understood.
  
  I translate proprie et stricte as “properly and strictly,” preserving both terms. Cottingham’s “in its strict and narrow sense” emphasizes that fewer thoughts fall within this use of “idea,” but “properly” directs attention to what qualifies a thought for that use. An idea of a lion is a strict idea because it internally represents a lion. The question is whether every other thought called an idea must likewise contain an internal representation of a thing. Descartes’s reply expressly distinguishes common notions from ideas of things properly so called. Understanding a principle such as nothing comes from nothing therefore cannot simply be treated as another case of representing a thing in the way that the idea of a lion does. The principle qualifies as an idea under the broader use, while its actual content must be examined to determine whether understanding it involves any represented thing. Preserving “properly” makes the reason for the distinction clearer. The broader category includes thoughts that need not qualify as ideas by satisfying the strict requirement of internally representing a thing.
  
  The repeated use of proprie connects the restriction of the universal statement directly to the treatment of common notions. Descartes first specifies that the statement concerns ideas taken proprie et stricte, then explains that ideas of common notions non sunt ideæ rerum proprie. My renderings, “properly and strictly” and “not properly ideas of things,” preserve the verbal connection. The qualification that identifies the ideas governed by the statement is the same qualification under which common notions are denied the status of ideas of things. Cottingham’s “strict and narrow” and “strictly speaking” make this repetition less visible. Retaining it enables the reader to follow Descartes’s answer without overlooking either part. The claim about ideas being as if of things concerns the proper and strict use, while Descartes’s reply nevertheless expressly recognizes common notions as ideas under the broader use. Common notions are thus acknowledged Cartesian ideas, not cases that an interpreter may discard merely because they are not properly ideas of things.
  
  The words autem hîc, which I render as “however” and “here,” identify how Descartes answers Burman’s proposed counterexample. Burman introduces the idea of nothing against the requirement that ideas be as if of things. Descartes’s reply calls that idea negative and scarcely an idea, then adds, in my translation, “the author, however, takes ‘idea’ here properly and strictly” (AT V 153; CSMK III 338). “However” marks the contrast between the idea Burman introduces and the strict ideas covered by the requirement. “Here” identifies the particular statement being defended, rather than declaring that Descartes always uses “idea” strictly. The reply therefore answers Burman by restricting the requirement’s application. It does not preserve an unrestricted requirement by turning Nothing into a thing represented by the idea. This matters because treating Nothing as a represented object would assume that the idea of nothing must contain a represented thing, although Burman introduces it precisely as an idea that is not of a thing. Descartes’s reply answers the objection by identifying the statement’s restricted scope. The requirement that ideas be as if of things applies to ideas understood properly and strictly. An idea of nothing, described in the reply as negative and scarcely an idea, therefore does not refute that requirement merely because it can be called an idea in a broader sense. If Descartes meant that everything he calls an idea must be of a thing, the idea of nothing would be a counterexample. But his reply limits that requirement to strict ideas, so a thought called an idea only in the broader sense does not contradict it.
  
  Finally, “but then ‘idea’ is taken more broadly” preserves the comparison expressed by latius (“more broadly”) by relating the use that includes common notions to the proper and strict use just identified. The word “idea” applies to common notions even though Descartes’s reply expressly denies that they are properly ideas of things. Understanding a principle such as nothing comes from nothing is a conscious thought, and its recognition as an idea does not require treating the principle itself as a represented thing comparable to the lion represented in an idea of a lion. Cottingham’s “a rather extended use” can suggest an unusual stretching of terminology because “extended” invites the reader to regard the application to common notions as a departure from the term’s usual application, while “rather” can reinforce that impression. My “taken more broadly” presents their inclusion directly as a recognized broader use of “idea.” Common notions consequently remain among the mental states that must be examined when assessing the claim that every Cartesian idea internally represents something. Excluding them from strict ideas does not exempt a defender of that universal claim from explaining what they allegedly represent. ↩︎
• 12
  Steven Nadler, “The Doctrine of Ideas,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 87. The surrounding discussion distinguishes broad and strict uses and illustrates the proposed strict core with desire for an ice cream cone. ↩︎
• 13
  Nadler, “The Doctrine of Ideas,” 91–92, especially 92. The defining-feature passage is on printed page 92 in the volume. It juxtaposes the strict classification at AT VII 37 with the claim about objective being by nature at AT VII 42. ↩︎
• 14
  Vere Chappell, “The Theory of Ideas,” in Essays on Descartes’ Meditations, edited by Amélie Oksenberg Rorty (Berkeley: University of California Press, 1986), 177–98, especially 178–79; Lawrence Nolan, “The Ontological Status of Cartesian Natures,” Pacific Philosophical Quarterly 78, no. 2 (1997): 169–94, especially 174–76. See Chapter One, Table 1.1, VC1–3 and LN1–5. ↩︎
• 15
  David Cunning, “Descartes on Sensations and Ideas of Sensations,” in An Anthology of Philosophical Studies (Athens: ATINER, 2006), 17–32, especially 18, 23, and 27. See Chapter One, Table 1.1, DC1–5. ↩︎
• 16
  John Carriero, Between Two Worlds: A Reading of Descartes’s Meditations (Princeton, NJ: Princeton University Press, 2009), 20–21, 158, 405, and 409. See Chapter One, Table 1.1, JC1–4. ↩︎
• 17
  Kurt Smith, “Descartes’s Ontology of Sensation,” Canadian Journal of Philosophy 35, no. 4 (2005): 563–84, at 563. Smith calls these claims an inconsistent triad and argues that Descartes accepts (1) and (3) but rejects (2). ↩︎
• 18
  Michael Della Rocca, “Judgment and Will,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 142–59, at 146. ↩︎
• 19
  Claudia Lorena García, “Descartes: Ideas and the Mark of the Mental,” History of Philosophy & Logical Analysis 3, no. 1 (2000): 21–53, especially 38 and 46. See Chapter One, Table 1.1, CLG1 and CLG5–6. Whether García’s account supports the weaker universal thesis depends on whether the representation she attributes to sensations is internal. ↩︎
• 20
  My translation, “by their own nature,” is philosophically preferable to Cottingham’s “by their very nature” as a translation of Descartes’s Latin: “iste modus essendi objectivus competit ideis ex ipsarum naturâ.” (AT VII 42; CSM II 29)
  
  The crucial expression is ex ipsarum naturâ. The emphatic pronoun ipsarum directs attention to the ideas themselves, whose nature grounds the objective mode of being. The preposition ex identifies that nature as the source or ground of the objective mode of being.
  
  There are three reasons to prefer “own”:
  
  (1) Grammatical precision. “Their own nature” makes the reference to the ideas themselves explicit. “Their very nature” is grammatically defensible, but it emphasizes the nature itself rather than its belonging to the ideas.
  
  (2) Philosophical precision. “Own” directs attention to what belongs to the ideas under discussion in virtue of their nature; I identify those ideas as strict ideas from the surrounding argument. It does not, by itself, establish that every mental mode qualifies as an idea possessing objective reality.
  
  (3) Interpretive significance. The universalist interpretation cannot derive its unrestricted scope merely from ipsarum. It must independently establish which cognitions fall within the extension of ideis in this passage.
  
  There is, however, one qualification. The Latin ipsarum can legitimately be translated as “very” when “very” functions as an intensifier of identity rather than degree. My translation makes the reference to the ideas themselves more explicit, but neither rendering determines whether Descartes means every thought he calls an idea or only the strict ideas whose represented contents he has been comparing. ↩︎
• 21
  Second Replies, Definition II. The definition continues across AT VII 160–61. English quotations follow CSM unless another translation is identified. ↩︎
• 22
  A later reflective thought takes that act of willing as its represented subject matter. Such reflection may contain objectively real content representing the earlier act. But this does not show that the original act of willing represented itself or that its volitional aspect consisted in objectively real representational content. ↩︎
• 23
  David Cunning excludes sensations from the category of ideas: “A Cartesian sensation is a mode of mind but not an idea. If it is representational, it is not representational in virtue of having objective reality but in virtue of something else.” (“Descartes on Sensations and Ideas of Sensations,” in An Anthology of Philosophical Studies, vol. I, edited by Patricia Hanna, Adrianne Leigh McEvoy, and Penelope Voutsina, Athens, Greece: Atiner Publishing, 2006, 18)
  
  Cunning later explains his restriction: “Another is that ideas are representational. Indeed, Descartes says that although something can be dubbed a ‘thought’ in a loose sense so long as it is a mode of mind, the term ‘idea’ is to be reserved for modes of mind that are of objects.” (23)
  
  This restriction fits the strict use of “idea” in the Third Meditation (AT VII 37; CSM II 25–26), but Descartes also recognizes a broader use. In the Fourth Replies, he expressly calls colour and cold sensory ideas while considering their failure to represent anything real (AT VII 234; CSM II 163).
  
  I agree with Cunning that secondary-quality sensations lack objective reality, but reject his exclusion of sensations from Cartesian ideas.
  
  Cunning also locates material falsity in an idea of a sensation (18), whereas I argue that a materially false sensation itself has a misleading presentational character before any false judgment is made. His exclusion of sensations therefore differs from my rejection of the universal theses, which concern every idea in Descartes’s broad sense. ↩︎
• 24
  Second Replies, Definition III. The initial ellipsis omits the introductory words of the definition. ↩︎
• 25
  The Third Meditation identifies thoughts that receive the name “idea” properly and strictly through their represented subject matter. (AT VII 37; CSM II 25) Definition II in the Second Replies explains an idea through immediate awareness of a thought. (AT VII 160–161; CSM II 113–114) The reply to Hobbes expressly includes volition and fear because they are immediately perceived. (AT VII 181; CSM II 127–128) The Fifth Replies also rejects restricting ideas to images in the imagination. (AT VII 366–367; CSM II 253)
  
  The inclusion of volition and fear establishes that the broad use covers the thought with its experienced volitional or affective character. It does not establish that the complete thought lacks represented subject matter. Fearing a lion includes objectively real content representing the lion, although the experienced fear is not itself another represented item. Nadler can therefore accept these passages while maintaining that every broad-sense idea contains strict representational content. The distinction between the two uses must be preserved, but neither the presence nor the absence of objective reality in every thought admitted under the broad use follows from that distinction alone. ↩︎
• 26
  Steven Nadler, “The Doctrine of Ideas,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 87. ↩︎
• 27
  Descartes distinguishes “a spontaneous impulse” to believe that ideas resemble external things from knowledge supplied by the natural light. He recalls that natural impulses have previously misled him. (AT VII 38–39; CSM II 26–27)
  
  He also allows that a faculty within himself, not yet fully understood, could produce ideas independently of his will without assistance from external things. Finally, even an external origin would not guarantee resemblance. His sensory and astronomical ideas of the sun differ, and he rejects the sensory appearance as an accurate likeness of the sun. (AT VII 39; CSM II 27) ↩︎
• 28
  I prefer “by their own nature” for ex ipsarum naturâ because ipsarum emphasizes the ideas themselves. The feminine plural pronoun refers to the ideas mentioned by ideis; its genitive case identifies whose nature Descartes means. Cottingham’s “by their very nature” is defensible, but “their own” makes this reference more direct. Both translations preserve the necessity of what belongs to the ideas by nature.
  
  The concluding ex earum naturâ, “by their nature,” refers to the causes of ideas. Descartes uses the ordinary earum here, whereas the earlier ipsarum is emphatic. Cottingham renders both phrases with “their very nature.” My translation retains the difference.
  
  I translate saltem primis & præcipuis as “at least the first and principal causes.” Primis identifies the first causes; præcipuis identifies the principal or chief causes. Descartes immediately explains why the causal account must reach them: an indefinite succession of ideas originating from other ideas would leave unexplained the source of the reality they contain objectively. The ultimate cause must possess sufficient reality formally or eminently. “Principal” makes this explanatory role clearer than Cottingham’s “most important,” although his wording is lexically defensible. Repeating “causes” also keeps that role explicit.
  
  The idea of God illustrates the issue. Its occurrence is a finite mental mode, but its content represents an infinite substance. The cause must account for that infinite objective reality. The same requirement applies according to the appropriate rank in other cases: an idea of a lion contains the objective reality of a finite substance.
  
  Finally, iste, “this,” points back to the objective mode of being just explained. Cottingham’s “the” leaves that reference implicit. Translating it as “this” makes the continuity of the argument more apparent, but the demonstrative does not by itself restrict the statement to strict ideas. Nor does ipsarum supply that restriction merely by emphasizing the ideas. My identification of them as strict ideas rests on Descartes’s comparison of represented contents and his explanation of the reality their causes must possess. (AT VII 42; CSM II 29) ↩︎
• 29
  My translation of Et quia nullæ ideæ nisi tanquam rerum esse possunt is “And because there can be no ideas except as if of things.” (AT VII 44; CSM II 30) I prefer “as if” to Cottingham’s “as it were” because the choice affects how the reader understands an idea’s being of something. “As it were” readily invites a qualified attribution, whereas “as if” foregrounds the difference between how something appears or behaves and whether it possesses the attributed character. Consider two descriptions of someone’s position in a department. Saying that he is, as it were, the department’s father credits him with a figurative paternal role. Saying that he behaves as if he were the department’s father describes his conduct without granting him that role. He might simply be assuming an authority he lacks. The distinction is between attributing a character in a qualified sense and describing something as if it possessed that character without making the attribution.
  
  Descartes’s discussion of bodily pain supplies a directly relevant example. In the Sixth Meditation, he writes “sentietur dolor tanquam in pede,” which Cottingham translates as “then pain will be felt as if it were in the foot.” (AT VII 88; CSM II 61) Cottingham here renders tanquam as “as if.” The surrounding explanation concerns a bodily disturbance occurring elsewhere along the nerves, or in the brain itself, that produces the same brain motion ordinarily caused by injury to the foot. The resulting pain is nevertheless felt in the foot. Its felt location belongs to the sensory experience itself. The person need not first experience an unlocated pain and then judge that it is in the foot. “As if” expresses this bodily location as experienced without affirming that the bodily disturbance occurs there.
  
  A cold sensation requires the same attention to what is already present in the experience. On my interpretation, phenomenal coolness is a formally real qualitative determination of the sensory mental mode, and it can be felt in the left hand rather than the right. Its being a mental determination does not prevent it from being experienced as bodily located. The sensation can therefore be of cold through its felt coolness and can present that coolness as located in the hand before any judgment is made. But coolness felt in the hand, the ice cube’s causing that experience, and the attribution of the same phenomenal coolness to the ice are different matters. Neither the felt bodily location nor the identification of the ice as the cause establishes that the ice possesses the qualitative character experienced. Nor should the sensory attribution of location simply be identified with an objectively real representation of the hand. Chapter Five examines the argument that felt bodily location requires such an internal representation.
  
  These examples explain the philosophical importance of translating tanquam rerum as “as if of things.” The qualification permits the experience’s own apparent bodily attribution to be acknowledged without treating that attribution as an established account of what the body possesses or what the sensation internally represents. By contrast, “as it were of things” can encourage a reader to assume that a thing is already represented and to qualify only the manner of its representation. For example, the sensation may be taken to contain an obscure representation of coolness in the ice, although the presence of that represented bodily quality is precisely what requires argument. My translation makes this step visible. Just as pain felt as if in the foot does not establish that its bodily cause is in the foot, coolness experienced as if belonging to a body does not establish that the body possesses phenomenal coolness. The force of tanquam is to keep the sensory appearance explicit, so that what is felt is not silently converted into an assertion about the body or an assumption about the sensation’s objectively real content. ↩︎
• 30
  My translation of the reported reply in the Conversation with Burman, 16 April 1648. This is a report of the exchange, rather than a passage from a work published by Descartes. The AT apparatus identifies Burman’s objection as directed at the statement that ideas can only be as if of things (AT VII 44; CSM II 30). Burman is not directly challenging the earlier statement that objective being belongs to ideas by their nature (AT VII 42; CSM II 29).
  
  Descartes’s reply therefore expressly restricts the statement about ideas being as if of things to strict ideas. It also supports my reading of the earlier statement by showing that Descartes’s apparently universal language about ideas does not automatically include every thought he calls an idea. The argument for restricting AT VII 42 comes from his comparison of represented contents and their causes, as developed in Section 2.5.
  
  My translation preserves several features of the Latin that explain how Descartes’s reported reply to Burman restricts the claim that ideas can only be as if of things. Burman challenges that claim with the idea of nothing. If every idea must be of a thing, an idea of nothing appears to be a counterexample. Descartes’s reply identifies the relevant use of “idea” as proprie et stricte, “properly and strictly,” and then acknowledges ideas of common notions that non sunt ideæ rerum proprie, “are not properly ideas of things.” He explains their inclusion by adding that idea latius sumitur, “‘idea’ is taken more broadly” (AT V 153; CSMK III 338). By distinguishing these uses of “idea,” Descartes limits the requirement to strict ideas while expressly recognizing common notions as ideas in the broader sense. Someone defending universal internal representation cannot therefore assume that every thought Descartes calls an idea must satisfy the requirement he attributes to ideas properly and strictly understood.
  
  I translate proprie et stricte as “properly and strictly,” preserving both terms. Cottingham’s “in its strict and narrow sense” emphasizes that fewer thoughts fall within this use of “idea,” but “properly” directs attention to what qualifies a thought for that use. An idea of a lion is a strict idea because it internally represents a lion. The question is whether every other thought called an idea must likewise contain an internal representation of a thing. Descartes’s reply expressly distinguishes common notions from ideas of things properly so called. Understanding a principle such as nothing comes from nothing therefore cannot simply be treated as another case of representing a thing in the way that the idea of a lion does. The principle qualifies as an idea under the broader use, while its actual content must be examined to determine whether understanding it involves any represented thing. Preserving “properly” makes the reason for the distinction clearer. The broader category includes thoughts that need not qualify as ideas by satisfying the strict requirement of internally representing a thing.
  
  The repeated use of proprie connects the restriction of the universal statement directly to the treatment of common notions. Descartes first specifies that the statement concerns ideas taken proprie et stricte, then explains that ideas of common notions non sunt ideæ rerum proprie. My renderings, “properly and strictly” and “not properly ideas of things,” preserve the verbal connection. The qualification that identifies the ideas governed by the statement is the same qualification under which common notions are denied the status of ideas of things. Cottingham’s “strict and narrow” and “strictly speaking” make this repetition less visible. Retaining it enables the reader to follow Descartes’s answer without overlooking either part. The claim about ideas being as if of things concerns the proper and strict use, while Descartes’s reply nevertheless expressly recognizes common notions as ideas under the broader use. Common notions are thus acknowledged Cartesian ideas, not cases that an interpreter may discard merely because they are not properly ideas of things.
  
  The words autem hîc, which I render as “however” and “here,” identify how Descartes answers Burman’s proposed counterexample. Burman introduces the idea of nothing against the requirement that ideas be as if of things. Descartes’s reply calls that idea negative and scarcely an idea, then adds, in my translation, “the author, however, takes ‘idea’ here properly and strictly” (AT V 153; CSMK III 338). “However” marks the contrast between the idea Burman introduces and the strict ideas covered by the requirement. “Here” identifies the particular statement being defended, rather than declaring that Descartes always uses “idea” strictly. The reply therefore answers Burman by restricting the requirement’s application. It does not preserve an unrestricted requirement by turning Nothing into a thing represented by the idea. This matters because treating Nothing as a represented object would assume that the idea of nothing must contain a represented thing, although Burman introduces it precisely as an idea that is not of a thing. Descartes’s reply answers the objection by identifying the statement’s restricted scope. The requirement that ideas be as if of things applies to ideas understood properly and strictly. An idea of nothing, described in the reply as negative and scarcely an idea, therefore does not refute that requirement merely because it can be called an idea in a broader sense. If Descartes meant that everything he calls an idea must be of a thing, the idea of nothing would be a counterexample. But his reply limits that requirement to strict ideas, so a thought called an idea only in the broader sense does not contradict it.
  
  Finally, “but then ‘idea’ is taken more broadly” preserves the comparison expressed by latius (“more broadly”) by relating the use that includes common notions to the proper and strict use just identified. The word “idea” applies to common notions even though Descartes’s reply expressly denies that they are properly ideas of things. Understanding a principle such as nothing comes from nothing is a conscious thought, and its recognition as an idea does not require treating the principle itself as a represented thing comparable to the lion represented in an idea of a lion. Cottingham’s “a rather extended use” can suggest an unusual stretching of terminology because “extended” invites the reader to regard the application to common notions as a departure from the term’s usual application, while “rather” can reinforce that impression. My “taken more broadly” presents their inclusion directly as a recognized broader use of “idea.” Common notions consequently remain among the mental states that must be examined when assessing the claim that every Cartesian idea internally represents something. Excluding them from strict ideas does not exempt a defender of that universal claim from explaining what they allegedly represent. ↩︎
• 31
  Steven Nadler, “The Doctrine of Ideas,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 87. The surrounding discussion distinguishes broad and strict uses and illustrates the proposed strict core with desire for an ice cream cone. ↩︎
• 32
  Nadler, “The Doctrine of Ideas,” 91–92, especially 92. The defining-feature passage is on printed page 92 in the volume. It juxtaposes the strict classification at AT VII 37 with the claim about objective being by nature at AT VII 42. ↩︎
• 33
  Vere Chappell, “The Theory of Ideas,” in Essays on Descartes’ Meditations, edited by Amélie Oksenberg Rorty (Berkeley: University of California Press, 1986), 177–98, especially 178–79; Lawrence Nolan, “The Ontological Status of Cartesian Natures,” Pacific Philosophical Quarterly 78, no. 2 (1997): 169–94, especially 174–76. See Chapter One, Table 1.1, VC1–3 and LN1–5. ↩︎
• 34
  David Cunning, “Descartes on Sensations and Ideas of Sensations,” in An Anthology of Philosophical Studies (Athens: ATINER, 2006), 17–32, especially 18, 23, and 27. See Chapter One, Table 1.1, DC1–5. ↩︎
• 35
  John Carriero, Between Two Worlds: A Reading of Descartes’s Meditations (Princeton, NJ: Princeton University Press, 2009), 20–21, 158, 405, and 409. See Chapter One, Table 1.1, JC1–4. ↩︎
• 36
  Kurt Smith, “Descartes’s Ontology of Sensation,” Canadian Journal of Philosophy 35, no. 4 (2005): 563–84, at 563. Smith calls these claims an inconsistent triad and argues that Descartes accepts (1) and (3) but rejects (2). ↩︎
• 37
  Michael Della Rocca, “Judgment and Will,” in The Blackwell Guide to Descartes’ Meditations, edited by Stephen Gaukroger (Malden, MA: Blackwell, 2006), 142–59, at 146. ↩︎
• 38
  Claudia Lorena García, “Descartes: Ideas and the Mark of the Mental,” History of Philosophy & Logical Analysis 3, no. 1 (2000): 21–53, especially 38 and 46. See Chapter One, Table 1.1, CLG1 and CLG5–6. Whether García’s account supports the weaker universal thesis depends on whether the representation she attributes to sensations is internal. ↩︎