Renaissance humanism is characterized by a revival of interest in ancient Greek and Latin writings. At the same time, the uses to which these ancient texts were put were typically contemporary, as Renaissance authors borrowed eclectically from their favorite ancients to make their modern claims. Montaigne wrote this way, but so did many others, many of whom did not feel obliged to cite their sources. For one thing, the audience they were writing for did not need to be informed of these allusions. These who knew their Cicero or Sextus did not require the pedantry of citation or quotation. Those who were unread in the classics could follow the argument all the same. And so Descartes could begin the first of his Meditations with allusions to well known Pyrrhonist arguments, without mentioning them by name.

Unacknowledged borrowings extended to other materials as well. This is especially evident in the use of translated materials not only from ancient Greek to modern Latin but also from modern Latin to the vernacular. Just as Gassendi translated Greek philosophy into the body of his Latin texts, so Gassendi's Latin was translated by English authors, without much effort being made to make this plain. Walter Charleton's Physiologia Epicuro-Gassendo-Chaltoniana (1654) was largely a translation from Gassendi's work, complete with Gassendi's borrowings from Epicurus. Since translated material could be expropriated this way, it is hardly surprising that philosophical themes and arguments would also make the rounds as commonplaces.

The transmission of Epicureanism into seventeenth-century England is a matter of importance. It is not merely that, as a matter of antiquarian interest, Boyle's corpuscularianism owes something to the ancient Epicureans in general.¹ It is rather that seventeenth-century science, which made such corpuscularianism plausible, is deeply indebted to what was understood at the time to be the empiricism of Epicurus. Establishing precise connections, beyond similarity of content, is an altogether different matter, since authors disdainful of citation were no more enthusiastic to confess the history of their own ideas. But here the expropriation by translation provides a useful tool in detecting the history of such transmission all the same, especially in the selection of lifted materials and their new employment.

It has recently been argued that “The Doctrine of Epicurus,” which is part XIII of Thomas Stanley's History of Philosophy (1660), was an important vehicle for this transmission, written as it was by a charter member of the Royal Society, an acquaintance of Locke and Boyle. Locke himself owned Stanley's History.² It is an even more compelling connection when we note that “The Doctrine of Epicurus” contained in Stanley's History is a literal translation of Gassendi's Philosophiae epicuri syntagma (1658), which in turn is largely a compilation of ancient texts that Gassendi had translated.

Gassendi actually listed many of his ancient sources in marginal notes alongside the columns of his text, while Stanley obligingly added more citations in his own notes, without finding it necessary to acknowledge Gassendi's role in their transmission.³ So this translation from ancient Greek to Gassendi's Latin to Stanley's English might well seem to be a path Epicurean empiricism had taken in getting from ancient Greece to modern England. Since the translations were competent, little seems to have been lost in their transmission, except of course for the control initiated by Gassendi's selectivity and presentation.

I. In that part of his Syntagma Philosophicum entitled De Logicae Fine (1658) Gassendi wrote: “It can be known certainly that Epicurus did not believe in only one criterion, that is, sense perception, but he called in two besides, namely ‘anticipation’ and ‘feeling,’ of which ‘anticipation’ pertains to the intellect.”⁴ Gassendi had arrived at this interpretation of Epicurus as a consequence of reading ancient sources. And Gassendi warns us that his documentation is the best he could provide, collected from Diogenes Laertius, Sextus Empiricus, and others, such as Cicero and Plutarch.⁵ The only primary texts available to Gassendi would have been the letters and sayings of Epicurus preserved in Book X of Diogenes Laertius. There was also the De rerum natura of Lucretius, whose manuscript had been rediscovered in the fifteenth century and printed in a notable edition by Lambinus in 1563. While Gassendi made much use of these primary materials in the course of his presentation, paraphrasing passages from Epicurus and Lucretius in the body of his text, he also adopted the overview of Epicureanism provided by Diogenes and to some extent attested to by Sextus.

Diogenes Laertius had stated, in a book which Gassendi had previously translated and annotated, that there were three Epicurean κριτηρια for settling disputes: αιsθηsειs, προληψειs, and παθη—that is to say, sense perceptions, pattern recognitions, as I translate προληψειs, and feelings.⁶ There is no explicit statement to this effect in the extant Epicurean corpus. In the Epicurean letters which accompany Diogenes' summation, Epicurus's numerous and formulaic references to the means for determining the truth are to sense perception and feelings alone.⁷ The word προληψιs does not occur in the canonical section of the Letter to Herodotus, where the sensory foundations of knowledge are referred to, although the term does occur later in the letter⁸ as well as in the Letter to Menoeceus, both times in an epistemic context.⁹ There is no explicit discussion of προληψιs in Lucretius, where the criteria for knowledge are presented uniformly as perceptual.¹⁰ Furthermore, παθη, or feelings, play a prominent evidentiary role in Epicurus's letters and also in Lucretius,¹¹ although Diogenes relegates them solely to the realm of ethics, leaving, according to Diogenes, only two effective criteria for determining the truth: perception and pattern recognition.¹²

Gassendi follows Diogenes' lead, accepting two epistemic criteria: perception and προληψιs, translating the latter as “anticipatio” or “praenotio” following Cicero,¹³ and relegating feelings to the realm of ethics.¹⁴ At the same time Gassendi gives considerable emphasis to προληψιs as a separate criterion of its own. In doing so Gassendi overrides the agreement of Cicero and Sextus Empiricus, for example, who insisted that the only measure of reality for the Epicureans was the senses.¹⁵ As Gassendi presents the Epicurean Canon, the pattern recognition of προληψιs is a separate act of intellection, not an act of sensation, thereby adding a rational criterion to a sensory one. In this way the criteria for knowledge contain a purely conceptual component in their very foundation, empiricism notwithstanding. Gassendi presents this as his interpretation of Epicurus.¹⁶ It is also his own view.

In the Syntagma philosophicum of 1658, a different Syntagma from the Philosophiae epicuri syntagma which Stanley used in his History of Philosophy, Gassendi presents an abbreviated version of that other work. Gassendi concedes that the Epicureans rejected the “logic” of the Stoics. Gassendi's concession is taken directly from Diogenes, who wrote that the Epicureans rejected “dialectic logic” as superfluous because they thought it sufficient for philosophers of nature just to make use of the sounds we make for things—τονs των πραγματων ϕθογγονs, echoing the words of the Letter to Herodotus.¹⁷

But unlike his ancient sources, Gassendi worries over the claim whether to forswear for the Epicureans any logic whatsoever, concluding that it must have been only Stoic dialectic which they had rejected.¹⁸ Gassendi examines the Epicurean foundations of knowledge as “dialectica” just the same, logic under a different name. Gassendi makes explicit what Diogenes assumes in practice, operating on the assumption that the Epicureans conducted philosophy along the same lines as their own contemporaries, the Stoics, had established—not of course exactly following in their footsteps but observing the general distinction between Logic, Natural Philosophy, and Ethics all the same—a distinction which Diogenes tells us the Stoics had invented.¹⁹ Following Diogenes, Gassendi could read the Epicureans as asking many of the same questions that the Stoics had addressed, despite their different answers. So Gassendi in his Philosophiae epicuri syntagma presents Epicurean doctrine according to the three Stoic divisions of philosophy, which Stanley dutifully repeats, “The First Part of Philosophy” being the Canonica.

In his abbreviated chapter on Epicurean logic, found in the Syntagma philosophicum, Gassendi follows up his discussion of the question of Epicurean logic with the observation that Epicurus first of all supposes that every question can be examined either de re or de voce.²⁰ This statement is ostensibly lifted from Diogenes' summation of Epicureanism: “Of inquiries, some concern things, others mere speech.”²¹ Gassendi adds that the former division is concerned with the way things are, while the latter is devoted to the meaning or significance we give to things. This seems to be Gassendi's interpretive intervention, for when he turns to the de voce question to provide the last two canons in his presentation of Epicurean dialectic, Gassendi acknowledges that he has gathered these together “ex fragmentis,” and what he has gathered together appear to be Epicurus's own remarks in the Letter to Herodotus, repeated by Lucretius, concerning the importance of straightforward language.²²

The use and abuse of words was a favorite Epicurean theme, but the sources of this abuse were always said to be the same: failing to index or label real facts or features in the world and as a consequence failing to provide any real reference for the words we use. When sounds we make fail to point to real things that underlie those utterances (των υποτεταγμενων ταιs ϕωναιs πραγματων),²³ they become noise instead of words. This is the point of Diogenes' division between de re and de voce, contrasting genuine inquiry into reality over against the empty sounds of bare speech (πeρι ψιλην την ϕωνην), which would afford nothing but an idle inquiry when there is nothing in the world for those empty sounds to name. Ignoring the force of Diogenes' ψιλην, Gassendi takes this latter sort of inquiry seriously, as the subject matter of Epicurean logic.

Gassendi takes Epicurus's reference to the things our utterances signal as a reference to the immediate signification we give to what we say, the thoughts or ideas in our minds first signified by speech. “Res subiectas vocibus, seu quid voces significent” is the way Gassendi translates the relevant passage in the Letter to Herodotus.²⁴ Gassendi interprets the things which underlie our utterances as the meanings of our terms, the notion or primary signification of our speech (vocis notionem, significationemve primariam), understanding Epicurus's Greek that way: το πρωτον ηννοημα καθ ηκαsτον ϕθογγον βλεπεsθαι.²⁵ Gassendi takes Epicurus's “empty utterances” (κενουs ϕθογγουs) to refer to meaningless utterances, sounds which lack thoughts or ideas to back them up.

Gassendi thereby finds a place in Epicurean theory for a purely conceptual criterion of meaning over and above the empiricism Gassendi's sources had insisted on. To defend this interpretation Gassendi draws upon a distinction between de re and de voce which he found in Book X of Diogenes Laertius, using this distinction to distinguish a theory of meaning from a theory of evidence. It is worth noting that this may not have been the distinction Diogenes intended. It is rather in Book VII of his Lives where Diogenes draws that sort of distinction, but there he attributes it to the Stoics, not the Epicureans.²⁶

II. In his Philosophiae epicuri syntagma, which Stanley appropriated for his History, Gassendi first explores the nature of truth before turning separately to the three criteria listed by Diogenes. Anticipating this discussion Gassendi establishes a distinction between the senses and the mind on the one hand, and our feelings on the other: “But forasmuch as natural things affect the Sense or Intellect, and moral things the Appetite or Will; for this reason, Criteria are to be taken from both these.”²⁷ This allows for a division between moral and epistemic criteria.

Gassendi then segregates the role of the mind, or intellect, from that of the senses:

From the Sense, nothing can be taken more than its Function, Sensation, which likewise is called Sense.

From the Intellect, forasmuch as besides the Function which it hath, whil'st like the Sense it contemplateth the thing, as if it were present and apparent, (whence the perception of a things appearing [perceptio apparentiae], which appeareth to be as well to the Intellect, as to the Sense, is called a Phantasie, or Appearance); forasmuch, I say, as besides this Function, it is proper to the Intellect to ratiocinate or discourse; there is therefore required a Praenotion or Anticipation, by looking upon which, something may be inferred.²⁸

The interpretation presented in these lines and dutifully translated by Stanley establishes the grounds for there being a separate, rational Epicurean criterion, in addition to a sensory one. Even if the mind could act as if it too were a sense organ in regarding the appearances presented before it, Gassendi argues that the mind must also manufacture concepts, or praenotions, as the vehicles for thought. Although this is presented as Epicurus's view, Gassendi in this instance is not quoting some ancient source but instead presenting his own interpretation of the philosophic necessity he believed Epicurus was responding to. It is also Gassendi's preferred view.

The Epicureans were notorious for saying that the mind could itself function as a sense organ, thereby accounting for dreams, mental visions, hallucinations, and the like, all of which might be thought to undermine the infallibility of our sense perceptions. To protect the authority of perception, the Epicureans insisted that there was nothing we could imagine or experience in this way which was not itself simply the impact of something from the outside, something we did recognize even though we might go on to make false claims about it.²⁹ Imagination was simply another form of direct perception attentive to external appearances, rather than something made up in the mind.

Gassendi acknowledges that doctrine by admitting that the mind could experience appearances too, and that all such appearances are true.³⁰ But Gassendi goes on to insist that because the Intellect can reason and speak, there must be required a separate conceptual criterion, separate from these imaging appearances. The specific content of this criterion Gassendi calls Praenotion or Anticipation, after the Epicurean προληψιs. This is the criterion for inquiries de voce and the foundation for what Gassendi takes to be the Epicurean theory of significance and meaning.

As far as the senses are concerned, Gassendi's presentation is accurate, that the senses of themselves make no claims on the world and are never false, that the senses do not interpret the world but instead present it to us, as it appears to be: “Now where there is a bare apprehension, not pronouncing any thing, there is no error or falsehood.”³¹ Gassendi is also quite clear that the appearances presented to the senses and the mind are not interpreted at all but are rather just the way the world looks as it presents itself before us, constructed as we are and as the world is.³² The sense organs provide purely mechanical transcriptions of the way the world looks to us to be.

Gassendi also appreciates how such a mechanical transcription limits the ability of the senses for getting at the truth, conceding that not all appearances are evident, even though all of them are true: “Evidence of Sense, I here call that kind of Sensation, or Appearance, which, all things obstructive to Judgment being removed, as distance, motion, indisposition of the Medium, and the like; cannot be contradicted. Whence to the Question, Whether a thing be such as it appears? We ought not to give a sudden Answer.”³³ But the source for Gassendi's statement here is taken from Cicero, in a passage where the Epicureans are being criticized from a Stoic point of view.

Gassendi understands the Epicurean contention, shared by the Stoics, that if there is to be any knowledge of reality at all, it is a knowledge that comes to us through the direct contact our senses have with the outside world, even though that knowledge may in turn be limited by the circumstances of perception. But it was Stoic philosophy which called attention to the limiting circumstances of perceiving, and Gassendi quotes them on this, in describing the Epicureans. It is not just that Gassendi employs Stoic categories of analysis found in ancient sources on Epicurus, but as he goes along Gassendi is apparently making improvements upon the Epicurean philosophy, improvements borrowed from the Stoics. Gassendi goes to explicitly Stoic sources and uses them to expound on Epicurus.

After examining the Epicurean sensory criterion, Gassendi next turns to the criterion of the intellect. Consider, for example, the first of Gassendi's four “Canons of Praenotion or Anticipation”:

(a) All Anticipation or Praenotion, which is in the Mind, depends on the Senses, either by Incursion, or Proportion, or Similitude, or Composition, I mean, that the Notion (or Idea, and Form as it were, which being anticipated is called Praenotion) is begotten in the Mind by Incursion (or Incidence,) when the thing incurreth into the Sense directly and by itself, as a man just before our eyes. By Proportion, when the Praenotion is amplified or extenuated. … By Similitude, when according to a thing first perceived by the Sense, we fancy another like it. … Lastly, by Composition, when we put as it were into one the distinct Notions which we have of two or more things, as when we so unite the Notions of a Horse and a Man, as that the Notion of a Centaur ariseth out them, but (b) not without some assistence of Ratiocination.³⁴

Stanley reorganizes the text by adding the markers (a) and (b) and the final italics, so that the final qualification is conjoined with the first, requiring that “Praenotion” explicitly depends on both the senses and ratiocination. As we have seen, Stanley's emphasis is perfectly consistent with Gassendi's own initial statement, that προληψeιs are separate acts of Intellection, though subsequent to perception and imagination.

The passage which Stanley translates from Gassendi is in turn taken from Diogenes Laertius, but it is not taken from Book X but rather from Book VII, where it explicitly describes the Stoic theory of ideas, and how those ideas are constructed in the intellect.³⁵ Once again, Gassendi goes to Stoic sources to explicate Epicurus. But this presents an immediate conflict with what Gassendi himself admits is the Epicurean doctrine of mental apparitions, where the appearance of a Centaur, say, is not for the Epicureans an intellectual composition invented by the mind but an externally constructed atomic ειδωλον, or physical image, impressed directly on the mind. Such “mental” presentations are for the Epicureans as rigidly segregated from mental constructions as αιsθηsιs is distinct from δοξα.³⁶

Gassendi identifies Epicurean προληψιs as if it were generically a kind of mental notion, εννοια, or idea. This is explicit doctrine of the Stoics, who typically regarded praenotions and general notions as alike in character, προληψειs and κοιναι εννοιαι, regarding them all as rational constructions, as opposed to solely something sensory.³⁷ And so the Stoics insisted that only creatures with an intellect could enjoy προληψειs too, along with the entire mosaic of mental creativity, while the Epicureans allowed προληψειs to deer, dogs, and parrots as well as generals, restricting rationality to matters of opinion.³⁸

In Book X of his discussion of Epicureanism Diogenes does discuss how Epicurean concepts, or επινοιαι, are generated from ϕαινομενα, and he alludes to this same fourfold process of construction which Diogenes had previously elaborated upon in his discussion of the Stoics, but here Diogenes insists that all such conceptual constructions arise just from sense perception (απο των αιsθηsεων γεγοναsι) with no mention made of ratiocination.³⁹ The coincidence of presentation between Books VII and X of Diogenes Laertius extends only to the generation of ideas and their composition and not to Epicurean προληψειs, except that Gassendi wants to understand Epicurean προληψειs as ideas, and so he completes the identification. Stanley emphasizes this interpretation in italics, that according to Gassendi the Epicurean intellectual criterion is effected by ratiocination and not just by sensation.

The Epicureans had argued that when the organism has sufficient experience to recognize the same appearance over and over again, it comes to recognize patterns in the world, and so it can recognize a horse when it sees one or a horse can recognize a man. And Lucretius reports that in this same way we can learn to recognize pestilence or peace. Epicurean προληψειs display habituated perceptions of ϕαινομενα extended over time.⁴⁰ Here there would be no need for ratiocination. Instead there is merely the reflexive recognition of repeated patterns of appearance, τουτεsτι μνημην του πολλακιs eξωθεν ϕανεντοs, as Diogenes put the point.⁴¹ As we have seen, Gassendi explicitly rejects this interpretation of προληψιs as appearance, in favor of the Stoic understanding of προληψιs as conceptual.

Gassendi took it upon himself to borrow Diogenes' account of the Stoics to explicate Epicurus. It served his own philosophical purposes to do so. This becomes evident as we examine Gassendi's second canon of interpretation concerning what Epicurean προληψιs is supposed to be: “Anticipation is the very Notion, and (as it were) Definition of the Thing; without which, we cannot Enquire, Doubt, Think, nor so much as Name any Thing.”⁴² That προληψιs is the sine qua non for enquiry, thought, and language is well attested from other sources besides Diogenes. The question is how it plays that role. This is the role Gassendi sees:

Hence it comes to pass, that, if it be demanded what any thing is, we define or describe it in such manner as it is, according to the Anticipation thereof which we have in our Mind. Neither do we thus only, being demanded what some singular thing is, as what Plato is, but also what an Universal is, as Man, not this or that, but considered in general; this is brought to pass according as the Mind, having seen many Singulars, and set apart their several Differences, formeth and imprinteth in herself the Anticipation of that which is common to them all, as an Universal Notion; reflecting upon which, we say, Man (for example) is something animate, and endued with such a Form.⁴³

According to Gassendi, Epicurean προληψειs provide the meanings of our general terms, and the role Intellection plays in the formation of προληψειs is that, upon the presentation of many single appearances, it sets apart their several differences. This is just one method of conceptual formation, the method of Incursion, but it is this method that establishes what our words for kinds or species signify.

Recognizing Plato when you see him may require an Anticipation too, but apprehension of a man as such, or anything requiring a general idea for its recognition, insists upon it. And since this Anticipation is a rational construction, it is open to explication by definition. In this way the conceptual definitions of our general ideas are established as the foundations for our knowledge. Now the Epicureans were famous for insisting that definitions were superfluous, since all that any organism needed was to see the patterns in the world and respond to them directly. Rather, it was the Stoics who insisted on definitions as the foundations for conceptual understanding. Once again, Gassendi prefers the Stoic point of view, even as an explication of Epicurus.

Gassendi also wants to insist, as the Stoics had before him, that these conceptual creations, mental contents as they are, are no less authoritative than the raw evidence of our senses:

But it is not necessary to confirm all things with exquisite Reasons or Arguments, and scrupulous forms of Reasoning, which are cried up by the Dialectics; For there is this difference betwixt an Argument and the Conclusion of the Reason, and between a slender Animadversion and an Admonition; that in one, some occult, and (as it were,) involved things are unfolded and opened; in the other, things ready and open are judged. But where there are such Anticipations as ought to be, then what will follow or not follow from them, or what agrees or disagrees with them, is perspicuously discerned, and naturally inferred, without any Artifice, or Dialectic Construction; wherefore we need only take care, that the Anticipation which we have of Things be clear and distinct.⁴⁴

Although they are rational representations responding to the senses, Anticipations are said to enjoy the authority of general truths because they are rationally perspicuous to the Intellect. These same Anticipations provide the vocabulary of thought and the general ideas “from which we infer Something, and thinking upon which we make Sumptions or Propositions, which are Maxims or Principles, by which that which is inferred or concluded is conceived to be demonstrated.”⁴⁵ So Gassendi argues on behalf of his Stoicized Epicureans that we suppose “the Anticipation of Vacuum” on the basis of “the Anticipation of a manifest thing” (Motion). Yet, in fact, for the Epicureans there could be no “Anticipation” of a vacuum, since a vacuum is something that we cannot directly experience.

The difficulty with Gassendi's interpretation is that not all Anticipations will then be true when they fail to be clear and distinct, when they are not “perspicuous and manifest,” when they are not such as they “ought to be.” Yet Epicurean προληψειs were supposed to be a constant criterion of truth in explicit contrast to whatever conceptions or opinions we might form consequent upon experience. Once these προληψειs are interpreted as intellectual creations, their authority comes into question. By converting Epicurean προληψειs over to their Stoic counterparts, Gassendi has undermined their infallibility.

Gassendi compares Epicurean προληψειs with general ideas. In doing so Anticipations become conceptual foundations for knowledge, the vocabulary for our propositions. The senses are relegated to mechanical harbingers of consequent mental apprehensions. Yet only some of these ensuing mental conceptions will be truly manifest and evident. In this way Gassendi's presentation of Epicureanism is modified by a Stoic epistemology still wedded to Epicurean atomism. To some extent Gassendi may have taken this path because he was led to do so by his sources. But there was another path he could have chosen, and that he did not choose it suggests his own commitment to this doctrine, which he himself defends. It must not be forgotten that Renaissance humanists made use of ancient sources for their own contemporary purposes.

III. Gassendi's four-part Institutio logica was published in 1658 as part of his Syntagma philosophicum. It is the initial portion of the Institutio logica concerning the origin and nature of ideas that is of interest to us here. Subsequent divisions addressing the nature of propositions, syllogistic argument, and rational signification all presuppose ideas as the vocabulary of thought. Gassendi employs a wide variety of ancient sources in formulating his general views. Yet it is Gassendi's understanding of the second of the Epicurean Canons of Criteria, which Gassendi had termed “the Canons of Praenotion or Anticipation,” that proves especially formative for his theory of ideas, although he does not mention Epicurus by name. Of course I am not claiming that Gassendi's Logic as a whole is singularly Epicurean, but I am suggesting that Gassendi's theory of ideas specifically owes a special debt to his understanding of Epicurus.

The first division of Gassendi's logic is entitled: “De simplici rerum imaginatione,” and Gassendi begins by explaining what this means:

We are here using the word “Imagination” for Thought, or that action of the Mind which terminates in an image of the thing thought [rei cogitatae imaginem] passing in front of the Mind. This must be noted because the expression is sometimes used for the imaginative faculty, which many people call also by the Greek word Phantasia and assign to the lower part of the Soul, which is common in Man and Brutes; for indeed, at this level, Brutes also imagine.⁴⁶

This is similar to a point Gassendi had made on behalf of the Epicureans, insisting as he had that Epicurean Praenotions and Anticipations were not part of animal imagination but the work of the intellect instead. In this way Gassendi separated man from brutes precisely at the point of conceiving in our minds what it is we are experiencing through our senses, exactly as the Stoics had.

Gassendi wants to insist that such a presence in the intellect of the thing one is thinking of has authority as something rationally apprehended: “that image which is passing before the mind, indeed is almost thrust before it when we think [veluti obiicitur].”⁴⁷ Gassendi argues that it makes no difference what we call this image—“Conceptio, Apprehensio, Intellectio, Imago, Idea, Species, Notio, Praenotio, Anticipatio, seu anticipata notio (prout nempe fuit prius acquisita); Conceptus, Phantasma”—since it is the very same ratiocination we are referring to.⁴⁸ But for the sake of convenience and of clarity Gassendi decides upon “Idea” as his term. These are the foundations of knowledge for Gassendi, what we might call simple ideas, or common notions and cognitive impression, as the Stoics had once called them: κοιναι εννοιαι and ϕανταsιακαταληπτικν.

Gassendi's separation of intellect from sense makes something of a mystery how one contributes to the other. If sensations and sensory images directly presented to the mind are not themselves ideas, it is hard to understand how the senses make their contribution to our understanding. Gassendi addresses this issue in the first three canons of his presentation. First of all, the simple concept of a thing requires repeated experience, enough to enable us even just to recognize a particular man when we see him again.⁴⁹ This requires repeated patterns of sensation. Secondly, the kinds of ideas we are capable of owe something to the kinds of sense organs we are equipped with. As Gassendi puts it, a man born blind cannot conceive of color, but how it is he does conceive of color once he has the sight to do so is left by Gassendi unexplained and inexplicable.⁵⁰

Finally, to those who object that some of our simple ideas are too incredible and fantastic to owe their origins to the senses, Gassendi responds with the very same fourfold construction of ideas which he had explicitly attributed to the Epicureans but had taken from the Stoics. Here Gassendi lays greater emphasis on the method of incursion or incidence (περιπτωsιs) as our means for acquiring our most ordinary ideas of horses, plants and flowers, the sun and moon, and men: “There are then Ideas which are said to cross over by impressions through the Senses, and are impressed upon the Mind, which are ideas of things which themselves strike upon the Senses.”⁵¹ Our other ideas are made up from these, constructed in those other ways. Consequently, Gassendi's separation of the intellect from sense must not be taken to resemble Descartes's separation of the mind from the body. For Gassendi, the intellect is entirely dependent on the senses for the origins of its ideas. It is just that the intellect interprets what the senses have presented to it.

Gassendi's Ideas are not mental objects of awareness, despite his describing them as images.⁵² Gassendi's Ideas do not veil our understanding of the world by substituting for it as the reference of our thought. Rather, they are acts of native intellection directed at the world, consequent upon the world's own effects upon us. It is not that Gassendi's Ideas come to take the place of things; instead such ideas are ways we have of making sense of the reality we experience with our senses. Ideas are vehicles of signification, which explains their role in formulating propositions. They are mental interpretants of the causes of our physiological experience. My idea of the sun is not the sun that so affects my eyes, but it is of that sun nevertheless, as my mind conceives it, responding to the sun's own physical impressions on my sense organs.

The source of this representation may be thrust upon the intellect, but its representation is an act of intellection all the same, something animals are quite incapable of. In this way, what Gassendi thought the Epicureans had meant by their προληψeιs become our first ideas and indeed the very foundations for all our knowledge, leaving to physical sensations the ancillary role as necessary but not sufficient prerequisites for rational recognition.⁵³ As a consequence of intellectualizing these acts of recognition, not all our simple ideas of intellection will prove true, as when we think of Pegasus for instance.⁵⁴ But our most ordinary ideas will be nothing more significant than rational recognitions of what it is that is effecting our bodily sensations.

Gassendi presents the fourth Canon of his logic, which he then elaborates upon as follows:

Every Idea which crosses over through the Senses is singular; it is the Mind which forms a general idea out of similar singular ones. Since all the things which are in the World and which are able to strike upon the Senses are singulars, like Socrates, Bucephalus, this stone, this piece of grass, and the other things one can point to with a finger (and of course there are not sufficient proper names to enable every individual thing to be designated), the ideas which cross over from them into the Mind and stick there can only be singular.⁵⁵

Even singular ideas, once they have made it into the mind, are ideas or anticipations, just as Gassendi said they were in his parallel discussion of Epicurean προληψιs. All the same, there is a difference between a singular idea and a general one. In discussing Epicurus, Gassendi had maintained that general Anticipations, or Universal Notions, are “brought to pass according as the Mind, having seen many Singulars, and set apart their several Differences, formeth and imprinteth in herself the Anticipation of that which is common to them all, as an Universal Notion.” Here Gassendi elaborates on this very same procedure, eliding the transition between single real things and our corresponding singular ideas just as he previously had done, leaving to the method of incursion the authority to effect the transition from singular impressions made upon our sense organs to simple recognitions of the specific objects causing them.

This elision is somewhat awkward, since the Epicureans, I take it, had insisted that individual appearances were presented to us through physical impressions made upon our sense organs and responded to directly by the material mechanism of the mind. There was not an additional epistemic transition from singular real things to our conceptual recognitions of them over and above the physical transition from object to the eye. By contrast Gassendi needs this second sort of transition, too, while relying on the first to do the work of both. Ideas strictly cannot cross over from the world to our intellect, since they are creatures of the intellect all along.

On the one hand, Gassendi wants to claim that ideas directly impress themselves upon us as they cross over from the world. On the other hand, Gassendi wants to insist that strictly speaking there are no ideas of sensation; there are only ideas of intellection consequent upon our physical sensations. The former claim was strict Epicurean doctrine, except that it was not a theory of ideas rather than a description of our physiology. The latter was established Stoic doctrine and vulnerable to well-known skeptical objections denying that the world need be anything like our ideas of it, even at the level of individual experiences. Epicurean doctrine was seen by ancient skeptics to be invulnerable to that specific criticism, and it was natural of Gassendi to embrace Epicurean mechanism. Yet Gassendi insisted upon the role of Stoic common notions, too, once he understood Epicurus along those same lines.

Although attentive observation and reliance on firsthand experience, in preference to what others say or write, can help perfect the ideas we do have, all our ideas are alike in character as rational representations of our experience.⁵⁶ And ideas can arise from conversation and written definitions just as they can from direct experience, the only difference being that ideas which arise by incidence, where a thing has made its impression on our senses, are really of that thing itself, whereas ideas which are formed another way are not so much of that very thing itself, but rather the accommodation of some other idea we have within us to signify the other thing in question.⁵⁷ Such an approach to knowledge trades off empirical exactitude for rational understanding, but it by no means makes all our ideas merely fanciful creations.

According to Gassendi there are two complementary methods for formulating general ideas out of singular ones (uno aggregando, alio abstrahendo).⁵⁸ The first involves discernment of collective similarity, the second sets aside individual differences by abstraction. The former method of collection might at first appear to be closer to the original intent of Epicurean προληψιs as a purely sensory apprehension of the mind (επιβολη τηs διανοιαs), where the mind becomes physically accustomed to sufficiently similar appearances over time to be able to recognize them as the kinds of things they are, in just the way a cat comes to recognize a mouse. The mind would collect and set aside similar experiences without applying to them any protocol of similitude of its own invention. Nature would do the sorting and the organism would be responsive to it.

Such a simple method of assimilation would be insufficient for Gassendi's purposes, since he requires that the method of collection be an intellectual operation employing some rationally articulated protocol, so that the mind can continue to take an active role in subsequently collecting these collections together, reaching ever higher levels of generality: in going from the general idea of “man” to that of “animal” or “substance,” for example.⁵⁹ Consequently, general ideas formulated in this way, just by being collected together by the mind, suffer varying degrees of perfection and completeness, the more neatly they portray what it is the separate singulars have in common (repraesentat purius id in quo singularia conveniunt). It is now the custom, Gassendi points out, to allow the idea of mankind to encompass Americans as well as Europeans. This example in turn suggests how the protocol of similitude is itself susceptible to change, due to new discoveries and differences in historical discernment.⁶⁰ In simply recognizing similarities and formulating general ideas to encompass them, the mind makes its own decisions and must answer for them to other minds.

The further method of abstraction operates upon such collections and gives them greater definition, as it goes through the ideas individually to determine what it is precisely that the members have in common, while setting aside their separate differences.⁶¹ In this way generality and universality are achieved by taking features away from singular ideas, resulting in the general idea of a man, for instance, who is neither young nor old nor middle-aged, something impossible to imagine but necessary for rational discrimination and definition.⁶² There seems to be this difference between the method of assimilation and the method of abstraction: the latter is a more self-conscious and articulating process, requiring increasing intervention by the mind as it maps out its own path of similarity and difference, rather than responding to a pattern it can recognize as more or less presented to it by the senses.

Abstraction is a method of selective definition conducted at any level of generality, as the mind identifies a thing for the kind of thing it is by rationally deciding upon the nature of that thing, fleshing out the definition of the matching notion. But this too is required of our idea of Socrates and its definition just as it is of “man.”⁶³ Furthermore, relations among ideas affect our individual understanding of them, be they singular or general.⁶⁴ At its very onset, each of our ideas, singular or general, is delineated by reason, whose definition of the thing can be articulated, be it Socrates or substance. The more rationally comprehensive such a classification scheme becomes, the more vulnerable it is to losing touch with nature's very own organization.⁶⁵

There are then various sources of vulnerability affecting our ideas that Gassendi recognizes. There is the question of our own rational satisfaction with our way of estimating our experience. There is also the need to isolate our ideas from temperament and prejudice and failure of articulation, so that our ideas remain cool and collected, measured responses to experience.⁶⁶ To defend against these vulnerabilities, Gassendi's theory of ideas requires a complex epistemology, one which must take into careful consideration failures of rationality over and above any lack of physiological experience, which itself amounts to a relatively minor factor in the epistemic scheme of things. Although Gassendi establishes this epistemology on the basis of experience, what he means by experience as an intellectualized reaction requires an elaborate defence and as a consequence is only capable of a moderate success and a modest knowledge. In this way Gassendi's epistemology is not beyond the reach of skepticism.

It is historically interesting in this respect that, although Gassendi would appear to present himself as following the path of the Epicureans, the epistemology of ideas that he practices is largely Stoic in its character and origins, and well known for its vulnerability to skepticism. Whereas for the Epicureans problems with our organs and the limits of their vision provided the primary limitations upon the reach of our experience, Gassendi's theory must take those physical limitations in stride as it delves into the psychological sources of our errors and the logic of such mistakes. Whereas for the Epicureans successful sensory apprehension of the way the world looks is guaranteed by the very mechanism of perception, for Gassendi the epistemic character of human experience is never beyond interpretation.

IV. A philosophical system that begins with ideas for its foundations in the end depends entirely upon the agreement or disagreement of those ideas with each other. The relation between conceptual schemata and the world those schemata are about might then appear to be problematic, if only temporarily so, requiring elaborated protocols for authenticating that relation. What it is those protocols would regulate is the particular way in which a coherent set of ideas conceives of reality so as to suit and signify the way the world really is.

By contrast, a philosophical system that depends on perceptual recognitions merely as acts of interaction between one physical reality and another is not much of an epistemology at all. On such a view, animals equipped with sensory organs mechanically discriminate among the causes of their physiological sensations, without the need to think about it. The mechanism of such a system presupposes a thorough-going materialism, to accommodate the connections between what affects those organisms and what those organisms then respond to. This sort of mechanism makes reason seem superfluous.

Proponents of a mechanistic physiology have always met with some resistance. What the horsefly sees as it flies around my head has, for example, been commonly described in epistemic rather than mechanical terms. And so we say without reflection that the horsefly sees my head, that it has some conception of my head, thanks to using its eyes the way it does. Although such an attribution to flies is somewhat gratuitous, over the centuries there has always been some recognized level of organic sophistication, in which it is readily agreed that the creature in question formulates organizing conceptions of its own in response to physiological stimulation. So we say that the dog recognizes his master, although we might not want to say that an amoeba sees the light, no matter how it might respond.

In the seventeenth century it was common for advocates of a mechanical philosophy, from Descartes and Gassendi to Robert Boyle, to fathom animal perception, such as how dogs recognize their masters, as a purely mechanical affair. This allowed for one sort of pattern recognition which was purely automatic, in contrast with another sort of recognition that was intellectually organized and complete with ideas and conceptions. Animals might respond to repeated patterns of experience, without requiring general ideas to do so, or any ideas at all. Yet such bestial pattern recognition contrasted sharply with intelligent perception. In the history of the period the former was just a form of mechanism, while the latter required the presence of a mind to respond mentally to the mechanics of the body.⁶⁷

Epicurean materialism made no separation between its theory of knowledge and its physics. Gassendi made a place for one, following Diogenes Laertius and the Stoics. We can understand how this might have seemed to Gassendi a charitable interpretation, to comprehend Epicurean προληψειs as part of human rationality. Otherwise, lacking any theory of conceptually organized experience to account for the ways in which we conceive of the world, the Epicurean senses would appear to establish only a causal link between the world and our organs without going on to establish our conceptions of the ways things are. If Epicurean προληψειs were just a part of the mechanism of physiological discrimination, it would appear to prove inadequate to formulate the foundations for our knowledge, once those foundations were required to be the rational representations of human beings and not beast machines. For this reason, Gassendi emphasizes that there must be two epistemic criteria for the Epicureans: sensation and intellection.

The scholarly Gassendi was an appreciative reader of the ancients and extended his appreciation especially to Epicurus, whose physics was of such interest to him but whose reduction of the mind to matter was at the time anathema. Borrowing heavily from the Stoics, Gassendi attributes to the Epicureans an additional, ratiocinative foundation for their knowledge, a theory of ideas with empirical authority. The generality of such ideas, as forms conceived in the mind and said of the world, provides for the universality of knowledge. And Gassendi adopts this same theory of ideas in his own Logic.

Mechanically speaking, our human bodies are in contact with the world that surrounds us. Epicurean atomism proved to Gassendi to be the most convenient way in which to describe the nature of that contact. As human beings, we are also in a position to have knowledge of that world as it brushes up against our senses whenever we experience it. Yet the nature of that experience must enable us not merely to respond to whatever is affecting us but to conceive of it as well. So, following his reading of Epicurus, Gassendi insisted upon the mind's own appreciation of that experience, as the mind intervenes to organize and recognize the world presented to it.

The theory that he came up with was flawed and fragile in character, since ideas themselves could not themselves be mechanically transmitted from the outside into us. The method of incursion for the formulation of our most basic ideas proved susceptible to skepticism exactly at the point of transition from the mechanism to the mind, for there was no reason to be sure that the mind's reactive interpretation of a pattern impressed upon the bodily sense organs would be true to nature's intrinsic order.

The philosophical construction of an empiricism resting on ideas is the story of seventeenth-century philosophy. Here I have only sketched out Gassendi's contribution to such a theory of empirical ideas by discussing its Hellenistic heritage. It is no coincidence that, given such a heritage, skepticism should be such a constant worry in the period. Yet the alternative was unthinkable; namely, pure mechanism stripped of all conceptions. Even though he was happy to criticize Descartes's mentalism from an Epicurean point of view, Gassendi's own theory of ideas was just as hostile to the mechanistic reduction of mind to matter.

Notes

1. See for example, C. T. Harrison: “Bacon, Hobbes, Boyle, and the Ancient Atomists,” Harvard Studies and Notes in Philology and Literature, 15 (1933), 191-218 and “The Ancient Atomists and English Literature of the Seventeenth Century,” Harvard Studies in Classical Philology, 45 (1934), 1-79; also R. H. Kargon, “Walter Charleton, Robert Boyle and the Acceptance of Epicurean Atomism in England,” Isis, 55 (1964), 184-92, and Atomism in England from Hariot to Newton (Oxford, 1966).

2. Cf. R. W. F. Kroll, “The Question of Locke's Relation to Gassendi,” JHI, 45 (1984), 346-52. Gassendi's 1658 edition of his Philosophiae epicuri syntagma was reprinted from his earlier Animadversiones in decimum librum Diogenis Laertii of 1649, which was in turn derived from his De vita et doctrina epicuri, composed between 1633 and 1645. All subsequent citations to Gassendi's Philosophiae epicuri syntagma will be to the 1658 Lyon edition, abbreviated as ES. Cf. H. Jones Pierre Gassendi's Institutio Logica (Assen, 1981), vii-lxviii; also B. Rochot, Les Travaux de Gassendi sur Épicure et sur l'atomisme 1619-1658 (Paris, 1944).

3. See Stanley's reference to Gassendi in his marginal note (a) in Part XIII, Chapter 2, Canon 1 of his History of Philosophy (London, 1701), 549. This will be the edition I shall be citing from, subsequently abbreviated as HP. Stanley's section on Epicurus was first published as the third volume of his History, in 1660.

4. This is a correction of an unfortunate error in C. B. Brush's translation in The Selected Works of Pierre Gassendi (New York, 1972), 317. The reference is to Gassendi Syntagma philosophicum, De logicae fine (liber alter) (Lyon, 1658), 76 (ch. 4), abbreviated as SP.

5. SP, 52, cf. ES, 1-2.

6. Cf. Diogenes Laertius [D.L.], X, 31.

7. See, for example, sections 38, 55, and 82 of the Letter to Herodotus as well as R.S., XXIV, preserved in section 147 of D.L., X.

8. D.L., X, 72.

9. D.L., X, 123-24, discussed at length in Cicero, De natura deorum, I.

10. Cf. Lucretius, IV, 469-521.

11. For the passages in Epicurean writings concerning the evidence of feelings see the discussion in D. K. Glidden: “Epicurus on Self-Perception,” American Philosophical Quarterly, 16 (1979), 297-306, and “Sensus and Sense-Perception in the De rerum natura,” California Studies in Classical Antiquity, 12 (1980), 155-81.

12. D.L., X, 32-34. In this context see Diogenes' explicit attribution of a twofold epistemic criterion to the Stoic Chrysippus at VII, 54: κριτηρια θηsιν ειναι αιsθηsιν και προληψιν.

13. Gassendi SP, 52-56, and ES, 4-10. Cf. Cicero, De natura deorum, I, 44.

14. Cf. SP, 52-53, and ES, 5.

15. Cf. Cicero, Academica, II, 142, with II, 19, 79, 82, which suggests that what gives authority to Epicurean παθη and προληψειs is the general infallibility of the senses. This becomes especially clear in Sextus's presentation of Epicurean epistemology at M, 7, 203-16, which may have been taken from Antiochus's Canonica, one of Cicero's sources.

16. Most scholars accept Gassendi's interpretation of Epicurus. See, for example, A. A. Long, “Aisthesis, Prolepsis and Linguistic Theory in Epicurus,” Bulletin of the Institute of Classical Studies, 18 (1971), 114-33; A. Manuwald, Die Prolepsislehre Epikurs (Bonn, 1972); G. Striker, “Kriterion tes aletheias,” Nachrichten der Akademie der Wissenschaften in Göttingen, II (1974), 47-110; E. Asmis, Epicurus' Scientific Method (Ithaca, 1984), 19-80. I do not: Cf. D. K. Glidden, “Epicurean Prolepsis,” Oxford Studies in Ancient Philosophy, 3 (1985), 175-217.

17. SP, 52, and D. L., X, 30-31.

18. SP, 52. The status of an Epicurean logic is not discussed in the earlier ES and consequently not translated by Stanley. Instead, ES simply cites the claim of Diogenes about there not being a third part of philosophy, dialectic, and leaves it at that (4), and that is what Stanley translates (548), even though Stanley follows Gassendi's presentation of Epicurean doctrine in terms of the three divisions of philosophy, treating Epicurean dialectic as a separate division.

19. D.L., VII, 39, attributes the invention of this trichotomy to the Stoic Zeno of Citium.

20. SP, 52-53; cf. ES, 4, Stanley, HP, 549.

21. D.L., X, 34.

22. SP, 52, 55; ES, 4, 10. See, for example, Letter to Herodotus at D.L., X, 37, discussed at length in Peri physeos, XXVIII, as well as in Lucretius, V, 1028-90. Cf. also Sextus, M, 8, 11-13, 258; also Plutarch, Adv. Col., 1119f-20a.

23. Cf. Letter to Herodotus (D.L., 37) with D.L., X, 31, and Sextus, P, 2, 211-12.

24. ES, 10.

25. Letter to Herodotus (D.L., X, 37). Gassendi understands το πρωτον ηννοημα as a conception of our consciousness rather than an act of recognition, what the mind conceives rather than what is deposited upon it by the senses. Asmis, op. cit., 19-34, largely accepts Gassendi's interpretation. But the Greek ηννοημα is a hapax legomenon in Epicurus and, for that matter, is found uncontestably only once in Aristotle. Yet it seems to mean the same in both authors, designating something brought to mind (i.e., noticed by the mind) as opposed to something conceived of in the mind of its own invention. So the charge to look at the principle ηννοημα in one's mind is not a request for introspection but a demanded act of attention, in this case directed at one's perception of the world, as I understand the passage. Cf. D.K. Glidden, “Epicurean Semantics,” in ΣUZΗΤΗΣΙΣ: Studi sull' epicureismo greco e romano offerti a Marcello Gigante (Naples, 1983), 185-226.

26. Cf. D.L., VII, 41-44, as for example 42: και το ορικον δε ομοιωs προs επιγνωsιν τηs αληθειαs δια γαρ των εννοιων τα πραγματα λαμβανεται.

27. Stanley translation (HP, 549) of ES, 5.

28. Ibid.

29. Epicurus's technical term for this direct mental perception was επιβολη τηs διανοιαs, as at Herodotus, 38, and R.S. XXIV. It is discussed at length by Lucretius, IV, 722-826, 962-1010, 1030-36; Cicero, De natura deorum, I, 49, 75-76, 82-83, 105-114; Diogenes of Oenoanda, frg. 6, new frgs. 5-6.

30. Cf. ES, 4, as well as chapter xviii of part 2 (de natura) of his presentation, translated by Stanley in HP, 589-90.

31. Stanley translation (HP, 550) of ES, 5.

32. Cf. D.K. Glidden, “The Epicurean Theory of Knowledge” (Princeton Univ. Ph.D. Diss., 1971); G. Striker, “Epicurus on the Truth of Sense Impressions,” Archiv für Geschichte der Philosophie, 59 (1977), 125-42; C. C. W. Taylor, “All Perceptions Are True,” in Doubt and Dogmatism, ed. M. Schofield, M. Burnyeat, and J. Barnes (Oxford, 1980), 105-24.

33. Stanley translation (HP, 551) of ES, 7. Cf. Cicero, Academica, II, 7.19.

34. Stanley translation (HP, 552) of ES, 8.

35. D.L., VII, 52-53. Cf. Cicero, Academica, II, 30, De finibus, III, 33; Sextus, M, 9.393, 11.250.

36. Cf. Lucretius, IV, 739-44, quoted by Gassendi in ES and translated by Stanley, HP, 589.

37. Cf. M. Frede, “Stoics and Skeptics on Clear and Distinct Impressions,” in The Skeptical Tradition, ed. M. Burnyeat (Berkeley, 1983), 65-94.

38. Cf. Lucretius, IV, 962-1036.

39. D.L., X, 32: The sentence referring to the construction of Epicurean επινοιαι may even be an interpolation, since it interrupts the flow of argument which resumes with the ensuing sentence.

40. Asmis's account, op. cit. 19-80, is compatible with my own account on this point: cf. “Epicurean Prolepsis,” cited in n. 16 above.

41. D.L., X, 33, although the value of this testimony is somewhat tainted by Diogenes's use of a string of well-known Stoic expressions with which to describe the Epicurean doctrine further. This particular phrase also has a long history, going back to ancient medicine and Aristotle.

42. Stanley translation (HP, 553) of ES, 8.

43. Ibid.

44. Stanley translation (HP, 553) of ES, 9.

45. Ibid.

46. Gassendi, Institutio logica (IS) Part 1, 3, from the Jones 1981 edition, cited in note 2 above, with some changes, from Jones's translation.

47. Ibid., 3-4.

48. Loc. cit.

49. Ibid., 4.

50. Loc. cit.

51. Ibid., 5.

52. Cf. O. R. Bloch, La Philosophie de Gassendi: Nominalisme, matérialisme et métaphysique (The Hague, 1971), 7-29, 77-147; F. Duchesneau, L'Empirisme de Locke (The Hague, 1973), 92-119. I find it more illuminating to compare Gassendi's theory of Ideas with the Hellenistic debates between the Skeptics and the Stoics. That debate was never a phenomenalist one, and it is the one Gassendi studied closely. Cf. L. S. Joy, Gassendi the Atomist (Cambridge, 1987), 165-74.

53. IS, 4-20.

54. Ibid., 4.

55. Ibid., 6.

56. Ibid., 11-12.

57. Ibid., 12.

58. Ibid., 6.

59. Ibid., 6-7, 10-11.

60. Ibid., 10-11.

61. Ibid., 6-11.

62. Ibid., 6, 12.

63. Ibid., 15-16.

64. Ibid., 18-19.

65. Ibid., 7-9, 16-19.

66. Ibid., 12-14.

67. See, for example, M. Boas, “The Establishment of the Mechanical Philosophy,” Osiris, 10 (1953), 413-541.