1. The Criteria of Truth Are The Sensations, Anticipations, and Feelings
Epicurus wrote extensively on what he called the "Canon of Truth" -- not a list of doctrines to be accepted on faith, but a description of the natural faculties by which human beings are equipped to judge what is true and what is false. Cicero reports that Torquatus described this Canon as having "fallen as it were from heaven" -- meaning it is a natural and reliable standard that anyone can use. The three criteria are: the sensations (aistheseis), which provide our direct contact with the physical world; the anticipations or preconceptions (prolepseis), which are the natural conceptual categories formed through repeated experience and stored in memory; and the feelings (pathe) -- pleasure and pain -- which guide our evaluation of what is good and bad.
The Canon functions like a ruler or straightedge: it is the instrument by which we measure and test proposed truths, not itself a list of conclusions. Its practical significance is profound. By anchoring all reasoning in these three natural faculties, Epicurus rules out several competing approaches to truth that were prevalent in his time and remain influential today: abstract dialectical logic disconnected from observation, pure introspection claiming access to ideal forms or essences, and divine revelation. Whatever cannot be confirmed by the sensations, anticipations, and feelings -- or that is explicitly contradicted by them -- is not to be accepted. Whatever is confirmed by them, or at least not contradicted, can be accepted as a reasonable working conclusion about the world.
It is worth being explicit about the third criterion, feelings (pathe), which is the most practically consequential and the most frequently overlooked in purely theoretical discussions of the Canon. Pleasure and pain are not merely subjective preferences or biological side-effects: they are nature's own testimony about what is beneficial and what is harmful to the living creature. Every animal -- including every human being -- experiences pleasure and pain from the moment of birth, before any teaching or cultural conditioning has had any effect. This makes feelings the most direct and universal evidence available to us about what genuinely matters. The feelings are thus criteria of truth in the practical domain: they tell us what is truly good and what is truly bad, just as the senses tell us what is truly there and anticipations tell us what category of thing we are dealing with. The three criteria together constitute an integrated epistemological system: sensation grounds empirical knowledge, preconceptions ground conceptual knowledge, and feelings ground evaluative knowledge.
If you are going to reject any sensation absolutely, and not distinguish opinions reliant on evidence yet awaited from what is already present through sensation, through feelings, and through every focusing of thought into an impression, you will confound all your other sensations with empty opinion and consequently reject the criterion in its entirety. And if you are going to treat as established both all the evidence yet awaited in your conjectural conceptions, and that which has failed to earn attestation, you will not exclude falsehood, so that you will have removed all debate and all discrimination between correct and incorrect. (tr. Long & Sedley)
First, that the words we use should have been first seen what each denotes, so that by referring to this we can judge what is awaited by opinions, or questions or problems, so that they are not left undecided for us as we go on proving things ad infinitum, or so that we do not have empty words. For the first conception corresponding to each word must be observed and need no further demonstration, if we are going to have something to which we can refer questions, problems, and opinions. (tr. Long & Sedley)
Moreover, one must suppose that it is when something from external objects enters into us that we not only see but also think of their shapes. (tr. Long & Sedley)
So we should pay heed to those feelings which are present in us, and to our sensations - universal sensations for universal matters, particular ones for particular matters - and to all self-evidence which is present by virtue of each of the discriminatory faculties. (tr. Long & Sedley)
(1) Now for what reason that image is seen beyond the mirror, I will tell. Certainly it appears thrust far back inside. Just so we see things which are in fact outside a doorway through doors when the door affords an unobstructed view through it out to the open and displays many outside things. (2) For that vision too is brought about by two separate airs. (tr. Long & Sedley)
Thus Epicurus, in the Canon ('Yardstick'), says that sensations, preconceptions and feelings are the criteria of truth. The Epicureans add the 'focusings of thought into an impression.' (tr. Long & Sedley)
[Above: Long & Sedley excerpt. Below: the same passage in Bailey's translation.]
Preconception, they [the Epicureans] say, is as it were a perception, or correct opinion, or conception, or universal 'stored notion' (i.e. memory), of that which has frequently become evident externally: e.g. 'Such and such a kind of thing is a man.' For as soon as the word 'man' is uttered, immediately its delineation also comes to mind by means of preconception, since the senses give the lead. Thus what primarily underlies each name is something self-evident. And what we inquire about we would not have inquired about if we had not had prior knowledge of it. For example: 'Is what's standing over there a horse or a cow?' For one must at some time have come to know the form of a horse and that of a cow by means of preconception. Nor would we have named something if we had not previously learned its delineation by means of preconception. Thus preconceptions are self-evident. And opinion depends on something prior and self-evident, which is our point of reference when we say, e.g., 'How do we know if this is a man?' (tr. Long & Sedley)
However, some of our school want to refine this doctrine, and deny that sensation should be left to be the criterion of what is good or what is bad. Instead, they maintain, it can be understood by the mind's reasoning both that pleasure is to be sought for its own sake and that pain is to be avoided for its own sake. To this end, they say that there is as it were a natural idea implanted in our souls. (trans. Sedley)
[Note: this passage (Sedley's own translation) reads as a continuation of the same speech shown in the Rackham/Reid clips already attached to the surrounding On Ends 1.29-32 entries -- the corresponding Reid page is also attached below for direct context.]
When therefore we say that pleasure is the goal we do not mean the pleasures of the profligate or the pleasures of consumption, as some believe, either from ignorance and disagreement or from deliberate misinterpretation, but rather the lack of pain in the body and disturbance in the soul. For it is not continuous drinkings and revellings, nor the satisfaction of lusts, nor the enjoyment of fish and other luxuries of the wealthy table, which produce a pleasant life, but sober reasoning, searching out the motives for all choice and avoidance, and banishing mere opinions, to which are due the greatest disturbance of the spirit. And for this cause we call pleasure the beginning and end of the blessed life. For we recognize pleasure as the first good innate in us, and from pleasure we begin every act of choice and avoidance, and to pleasure we return again, using the feeling as the standard by which we judge every good. (Bailey)
[This entry combines two passages Bailey presents separately: the "profligate" disclaimer and the "beginning and end of the blessed life" formula. Below: both source pages.]
- Page scan: Epicurus, Letter to Menoeceus 128 (tr. Bailey), "beginning and end of the blessed life"
- Page scan: Epicurus, Letter to Menoeceus 129-131 (tr. Bailey), "profligate" disclaimer
- Online text: Epicurus, Letter to Menoeceus 128-9 (Feelings as the Starting Point of All Choice) -- Perseus Digital Library (Diogenes Laertius X, tr. Hicks, Loeb)
The internal sensations they say are two, pleasure and pain, which occur to every living creature, and the one is akin to nature and the other alien: by means of these two choice and avoidance are determined. Of inquiries some are concerned with things and some with nothing but words. (Bailey)
[A doxographical testimony (Aëtius's Placita, Book 4, on Psychology, in the chapter asking whether sensations and impressions are true) classifying pleasure and pain, according to Epicurus, among the "sense objects" themselves -- independent confirmation, from outside the Epicurean tradition's own reports, that the feelings function as empirical, sensory data within the Canon, on the same epistemic footing as sensation, rather than as some separate subjective add-on.]
Epicurus (says) that the pleasures and pains actually belong with the sense objects.
Since man has nothing left if sensations are removed from him, it must be the case that nature itself judges what is in accordance with or contrary to nature. What does it perceive or what does it judge except pleasure and pain, as a basis for its pursuit or avoidance of anything? (tr. Long & Sedley)
[Epicurus On The End] 'For my part I cannot conceive of anything as the good if I remove the pleasures perceived by means of taste and sex and listening to music, and the pleasant motions felt by the eyes through beautiful sights, or any other pleasures which some sensation generates in a man as a whole. Certainly it is impossible to say that mental delight is the only good. For a delighted mind, as I understand it, consists in the expectation of all the things I just mentioned - to be of a nature able to acquire them without pain... ' A little later he adds: 'I have often asked men who were called wise what they could retain as the content of goods if they removed those things, unless they wanted to pour out empty words. I could learn nothing from them; and if they want to babble on about virtues and wisdoms, they will be speaking of nothing except the way in which those pleasures I mentioned are produced.' (tr. Long & Sedley)
[Above: Long & Sedley excerpt. Below: the same passage in full surrounding context, tr. Yonge, Tusculan Disputations III (chapter XVIII).]
The Epicureans start off with Logic, for they expound "Canonics" first, treating of things evident and non-evident and allied matters. (Bury)
[Compare Against the Logicians II.330, where "those of the Epicurean School" press the Skeptic: "Either you understand what proof is, or you do not understand; and if you understand and have a notion of it, proof exists; but if you do not understand, how do you inquire into a thing of which you have no understanding at all?"]
Since, according to the sage Epicurus, it is not possible either to inquire or to doubt without a preconception, it will be well first of all to consider what "Grammatic" (or the Art of Grammar) is. (Bury)
[Sextus does not name Epicurus here, but the doctrine is the Epicurean one reported at Diogenes Laertius 10.32: all notions arise from the senses by direct contact, analogy, resemblance, and composition. Sextus repeats it at Against the Ethicists 250.]
In general, too, everything which is conceived is conceived either through the presentation of things manifest to the senses or through transition from things manifest, and this again in various ways -- at one time through resemblance, at another through composition, at another by analogy, and this again by way either of increase or of decrease. Thus it is through the presentation of things manifest that things like white and black and sweet and bitter are conceived; for these things, though they are sensible, are none the less conceived. And things are conceived through transition from things manifest either by way of resemblance (as, for instance, the absent Socrates from a likeness of Socrates); or by way of composition (as, for instance, from man and horse the Hippocentaur, which is neither man nor horse but compounded of both); or by analogy, which may either magnify or diminish the object. (Bury)
2. Language
In the ancient world it was common to hold that language was given to humanity by the gods and thus words have a supernatural basis. Epicurus rejected this contention and held that language, like everything else, develops naturally and without influence from supernatural forces.
This matters philosophically because the natural origin of language is directly connected to the Canon of Truth. If words refer to things apprehended by the senses -- because language developed as human beings named what they perceived -- then the meanings of words are anchored in real sensory experience rather than in Platonic forms, divine essences, or arbitrary conventions. The preconception attached to each word is the accumulated sensory experience of the things that word names. This makes language a tool for pointing to real things, not a separate realm of abstract meanings floating free of physical reality. When Epicurus insists that we must begin by being clear about what words mean -- so that we have something to refer to when we make claims or raise questions -- he is insisting that words be kept tethered to the sensory reality from which they arose. A word that has lost its connection to sensory experience becomes empty and generates confusion rather than understanding. This is one of the reasons Epicurus distrusted the elaborate technical vocabulary of dialectical philosophy: it had largely lost its sensory anchorage and was generating mere words rather than genuine knowledge.
We must take it that even nature was educated and constrained in many different ways by actual states of affairs, and that its lessons were later made more accurate, and augmented with new discoveries by reason -- faster among some people, slower among others, and in some ages and eras, owing to <individual needs, by greater leaps>, in others by smaller leaps. Thus names too did not originally come into being by coining, but men's own natures underwent feelings and received impressions which varied peculiarly from tribe to tribe, and each of the individual feelings and impressions caused them to exhale breath peculiarly, according also to the racial differences from place to place. Later, particular coinings were made by consensus within the individual races, so as to make the designations less ambiguous and more concisely expressed. Also, the men who shared knowledge introduced certain unseen entities, and brought words for them into usage. Hence some men gave utterance under compulsion, and others chose words rationally, and it is thus, as far as the principal cause is concerned, that they achieved self-expression. (tr. Long & Sedley)
- Page scan: Epicurus, Letter to Herodotus 75 (tr. Bailey) -- opening of sec. 75 shown (same page also carries secs. 73-74); continues onto the next page
- Page scan: Epicurus, Letter to Herodotus 76 (tr. Bailey), continuation, same page also carries secs. 77-78
- Online text: Epicurus, Letter to Herodotus 75-76 -- Perseus Digital Library (Diogenes Laertius X, tr. Hicks, Loeb)
Supposing that in those days we thought and said something equivalent, in the terminology which we then employed, to saying that all human error is exclusively of the form that arises in relation to preconceptions and appearances because of the multifarious conventions of language... (tr. Long & Sedley)
But perhaps this is not the moment to prolong the discussion by citing these cases? Quite so, Metrodorus. For I expect you could cite many cases, from your own past observations, of certain people taking words in various ridiculous senses, and indeed in any sense rather than their actual linguistic meanings. Whereas our own usage does not flout linguistic convention, nor do we alter names with regard to things evident. (tr. Long & Sedley)
It was nature that compelled the utterance of the various noises of the tongue, and usefulness that forged them into the names of things. It was rather in the way that children's inarticulacy itself seems to impel them to use gestures, when it causes them to point out with a finger what things are present. For everyone can feel the extent to which he can use his powers. The calf angrily butts and charges with his incipient horns before they have even protruded from his forehead. Panther and lion cubs already fight with claws, paws and biting at an age when their teeth and claws have barely appeared. Also, we see all birds putting trust in their wings and seeking the fluttering aid of their feathers. So to think that someone in those days assigned names to things, and that that is how men iearned their first words, is crazy. Why should he have been able to indicate all things with sounds, and to utter the various noises of the tongue, yet others be supposed not to have had that ability at the same time? Besides, if others had not already used sounds to each other, how did he get the preconception of their usefulness implanted in him? How did he get the initial capacity to know and see with his mind what he wanted to do? Again, one person could not subdue many and compel them to want to learn the names of things. Nor is it easy to find a way of teaching and persuading a deaf audience of what needs to be done: they would utterly refuse to tolerate any further his drumming into their ears the unfamiliar sounds of his voice. Lastly, why is it so surprising that the human race, with its powers of voice and tongue, should have indicated each thing with a different sound to correspond to a different sensation? After all, dumb animals, tame and wild alike, regularly emit different sounds when afraid, when in pain, and when happiness comes over them. You can appreciate this from plain facts. When the great flabby jaws of Molossian hounds start to snarl in anger, baring their hard teeth, their threatening noise when drawn back in rage is far different from when they fill the place with their barking. When, on the other hand, they try coaxingly to lick their pups with their tongue, or when they toss them with their paws and, powerful biters though they are, hold back their teeth and make play of nibbling them gently, they fondle them with a yelping noise quite different from when they whimper, cowering from a blow. If, then, different sensations compel animals, dumb though they are, to emit different sounds, how much more likely it is that mortal men at that time were able to indicate different things with different sounds. (tr. Long & Sedley)
As for the words, I mean the nouns and verbs, of which the men who sprang from the earth made the first utterance, let us not adopt Hermes as our teacher, as some say he was: that is manifest nonsense. Nor should we believe those philosophers who say that it was by coining and teaching that names were assigned to things, in order that men might have signs to facilitate their communication with each other. For it is absurd, indeed absurder than any absurdity, not to mention impossible, that someone should all on his own have assembled all those multitudes. There were no rulers at that time, and certainly no letters, seeing that there were no words, since it was about these <that they were meeting, so that it was not> by edict that their assembly was brought about - and having assembled them, instructed them like a schoolteacher, holding a rod, and touching each thing have said 'Let this be called "stone", this "stick", this " man", or "dog." (tr. Long & Sedley)
- Page scan: Diogenes of Oinoanda, Fr. 12 (tr. Smith)
Epicurus says that names are clearer than definitions, and that indeed it would be absurd if instead of saying 'Hello Socrates' one were to say 'hello rational mortal animal'. (tr. Long & Sedley)
[Above (following line): Long & Sedley excerpt. This clip: the same doctrine in Bailey's full translation, tr. Bailey.]
They [the Epicureans] reject dialectic as superfluous, saying that it is sufficient that natural philosophers should proceed in accordance with the word s belonging to things. (tr. Long & Sedley)
[Above: Long & Sedley excerpt. Below: the same passage in Bailey's translation.]
[The Epicureans say that] of inquiries, some are about things, others about mere utterance. (tr. Long & Sedley)
For if we are going to explain the words known to everybody, we would have to expound either all or some. But to expound all is impossible, whereas to expound some is pointless. For we will explain them either through familiar locutions or through unfamiliar. But unfamiliar words seem unsuited to the task, the accepted principle being to explain less known things by means of better known things; and familiar words, by being on a par with them, will be uninformative for illuminating language, as Epicurus says. For the informativeness of language is characteristically ruined when it is bewitched by an account, as if by a homoeopathic drug. (tr. Long & Sedley)
In the other branch of philosophy, logic, which concerns inquiry and argument, your master [Epicurus] seems to me unarmed and naked. He abolishes definitions. He teaches nothing about division and partition. He gives no advice on framing a deductive argument. He does not show how to solve sophisms or to distinguish ambiguous terms. (tr. Long & Sedley)
[Above: Long & Sedley excerpt. Below: the same passage in full surrounding context, tr. Reid, On Ends I.22 -- Reid renders it "utterly defenceless and without weapons" rather than "unarmed and naked."]
- Page scan: Cicero, On Ends 1.22 (tr. Reid)
Who is more in error than you [the Epicureans] about language? You completely abolish the class of sayables, to which discourse owes its existence, leaving only word s and name-bearers, and denying the very existence of the intermediate states of affairs signified, by means of which learning, teaching, preconceptions, thoughts, impulses and assents come about. (tr. Long & Sedley)
[Above: tr. Goodwin (1874), Against Colotes the Epicurean, full surrounding context. Below: the same passage as excerpted by Long & Sedley.]