(4th tetralogy : The Soul - 2nd dialogue of trilogy)

The table below compares the translation of a number of important Greek words from the analogy of the line (Republic, VI, 509c5-511e5) by different translators, in all their occurrences in the relevant section (as well as in the section Republic, VII, 533e7-534a8, which more or less recalls this section; references from this section appear in green in the table). A (G) after the reference and the ochre text indicate that this occurrence is spoken by Glaucon. When there is an entry for the word in my Lexicon of Greek words important for understanding Plato, the word in column 1 constitutes a link to this entry. The primary purpose of this table is not to compare the worth of the different translations, but to stress the difficulty of translating Plato and the danger for the reader of relying only on one translation, no matter how good it may be.

The complete Greek text of the two sections analyzed here is provided before the table in Latin transliteration, to allow those who are interested to see the words in their context. The words referenced in the table are in red in the Greek text, followed by a sequential number in parentheses that is a link to the row in the table corresponding to that occurrence of the word. In the table, the different occurrences of the same word have been grouped one after the other alternatively on a white or light grey background (except for the occurrences of the word eidos, which are on a light pink background) and the words are not listed in alphabetical order but rather by theme, and the words deriving from the same root are listed one after the other. And to help you find your way around the table, the list below presents the Greek words that appear in it in the alphabetical order of their transliteration into the Latin alphabet, each word in the list being a clickable link to the part of the table where it appears:

Republic, VI, 509c5-511e5

[509d] Noèson (1) toinun, èn d' egô, hôsper legomen, duo autô einai, kai basileuein to men noètou (2) genous (3) te kai topou (4), to d' au horatou (5), hina mè ouranou eipôn doxô soi sophizesthai peri to onoma. All' oun echeis tauta ditta eidè (6), horaton (7), noèton (8);
Echô.
Hôsper toinun grammèn dicha tetmèmenèn labôn anisa tmèmata, palin temne hekateron to tmèma ana ton auton logon (9), to te tou horômenou (10) genous (11) kai to tou nooumenou (12), kai soi estai saphèneiai kai asapheiai pros allèla en men tôi horômenôi (13) [509e] to men heteron tmèma eikones (14)--legô de tas eikonas (15) prôton [510a] men tas skias, epeita ta en tois hudasi phantasmata kai en tois hosa pukna te kai leia kai phana sunestèken, kai pan to toiouton, ei katanoeis (16).
Alla katanoô (17).
To toinun heteron tithei hôi touto eoiken, ta te peri hèmas zôis kai pan to phuteuton kai to skeuaston holon genos (18).
Tithèmi, ephè.
È kai ethelois an auto phanai, èn d' egô, dièirèsthai alètheiai te kai mè, hôs to doxaston (19) pros to gnôston (20), houtô to homoiôthen pros to hôi hômoiôthè;
[510b] Egôg', ephè, kai mala.
Skopei dè au kai tèn tou noètou (21) tomèn hèi tmèteon.
Pèi;
Hèi to men autou tois tote mimètheisin hôs eikosin (22) chrômenè psuchè (23) zètein anankazetai ex hupotheseôn (24), ouk ep' archèn (25) poreuomenè all' epi teleutèn (26), to d' au heteron--to ep' archèn (27) anupotheton (28)--ex hupotheseôs (29) iousa kai aneu tôn peri ekeino eikonôn (30), autois eidesi (31) di' autôn tèn methodon (32) poioumenè.
Taut', ephè, ha legeis, ouch hikanôs emathon (33).
[510c] All' authis, èn d' egô: rhaion gar toutôn proeirèmenôn mathèsèi (34). Oimai gar se eidenai (35) hoti hoi peri tas geômetrias te kai logismous (36) kai ta toiauta pragmateuomenoi, hupothemenoi (37) to te peritton kai to artion kai ta schèmata (38) kai gôniôn tritta eidè (39) kai alla toutôn adelpha kath' hekastèn methodon (40), tauta men hôs eidotes (41), poièsamenoi hupotheseis (42) auta, oudena logon oute hautois oute allois eti axiousi peri autôn didonai (43) [510d] hôs panti phanerôn, ek toutôn d' archomenoi (44) ta loipa èdè diexiontes teleutôsin (45) homologoumenôs (46) epi touto hou an epi skepsin hormèsôsi.
Panu men oun, ephè, touto ge oida (47).
Oukoun kai hoti tois horômenois (48) eidesi (49) proschrôntai kai tous logous peri autôn poiountai (50), ou peri toutôn dianooumenoi (51), all' ekeinôn peri hois tauta eoike, tou tetragônou autou heneka tous logous poioumenoi (52) kai diametrou autès, all' ou [510e] tautès hèn graphousin, kai talla houtôs, auta men tauta ha plattousin te kai graphousin, hôn kai skiai kai en hudasin eikones (53) eisin, toutois men hôs eikosin (54) au chrômenoi, zètountes [511a] de auta ekeina idein (55) ha ouk an allôs idoi (56) tis è tèi dianoiai (57).
Alèthè, ephè, legeis.
Touto toinun noèton (58) men to eidos (59) elegon, hupothesesi (60) d' anankazomenèn psuchèn(61) chrèsthai peri tèn zètèsin autou, ouk ep' archèn (62) iousan, hôs ou dunamenèn tôn hupotheseôn (63) anôterô ekbainein, eikosi (64) de chrômenèn autois tois hupo tôn katô apeikastheisin kai ekeinois pros ekeina hôs enargesi dedoxasmenois te kai tetimèmenois.
[511b] Manthanô (65), ephè, hoti to hupo tais geômetriais te kai tais tautès adelphais technais (66) legeis.
To toinun heteron manthane (67) tmèma tou noètou (68) legonta me touto hou autos ho logos (69) haptetai tèi tou dialegesthai (70) dunamei, tas hupotheseis (71) poioumenos ouk archas (72) alla tôi onti hupotheseis (73), hoion epibaseis te kai hormas, hina mechri tou anupothetou (74) epi tèn tou pantos archèn (75) iôn, hapsamenos autès, palin au echomenos tôn ekeinès echomenôn, houtôs epi teleutèn (76) katabainèi, [511c] aisthètôi pantapasin oudeni proschrômenos, all' eidesin (77) autois di' autôn eis auta, kai teleutai (78) eis eidè (79).
Manthanô (80), ephè, hikanôs men ou--dokeis gar moi suchnon ergon legein--hoti mentoi boulei diorizein saphesteron einai to hupo tès tou dialegesthai (81) epistèmès (82) tou ontos te kai noètou (83) theôroumenon è to hupo tôn technôn (84) kaloumenôn, hais hai hupotheseis (85) archai (86) kai dianoiai (87) men anankazontai alla mè aisthèsesin auta theasthai hoi theômenoi, dia de to mè ep' archèn (88) [511d] anelthontes skopein all' ex hupotheseôn (89), noûn (90) ouk ischein peri auta dokousi soi, kaitoi noètôn ontôn (91) meta archès (92). Dianoian (93) de kalein moi dokeis tèn tôn geômetrikôn te kai tèn tôn toioutôn hexin (94) all' ou noûn (95), hôs metaxu ti doxès (96) te kai noû (97) tèn dianoian (98) ousan.
Hikanôtata, èn d' egô, apedexô. Kai moi epi tois tettarsi tmèmasi tettara tauta pathèmata (99) en tèi psuchèi (100) gignomena labe, noèsin (101) men epi tôi anôtatô, dianoian (102) [511e] de epi tôi deuterôi, tôi tritôi de pistin (103) apodos kai tôi teleutaiôi (104) eikasian (105), kai taxon auta ana logon (106), hôsper eph' hois estin alètheias metechei, houtô tauta saphèneias hègèsamenos metechein.
Manthanô (107), ephè, kai sunchôrô kai tattô hôs legeis.

Republic, VII, 533e7-534a8

[533e] Arkesei oun, èn d' egô, hôsper to proteron, tèn men prôtèn moiran (108) epistèmèn (109) kalein, deuteran de dianoian (110), tritèn [534a] de pistin (111) kai eikasian (112) tetartèn: kai sunamphotera men tauta doxan (113), sunamphotera d' ekeina noèsin (114): kai doxan (115) men peri genesin (116), noèsin (117) de peri ousian (118): kai hoti ousia (119) pros genesin (120), noèsin (121) pros doxan (122), kai hoti noèsis (123) pros doxan (124), epistèmèn (125) pros pistin (126) kai dianoian (127) pros eikasian (128): tèn d' eph' hois tauta analogian (129) kai diairesin dichèi hekaterou, doxastou (130) te kai noètou (131), eômen, ô Glaukôn, hina mè hèmas pollaplasiôn logôn (132) emplèsèi è hosôn hoi parelèluthotes.

As might be expected, we see that the same Greek word can be translated by different English words (for instance, noûs is variously translated by "reason", "understanding" or "intelligence"), not only from one translator to another, but even by the same translator from one occurrence to another, depending on the context (for instance, the three occurrences of genos in the analogy are translated by Shorey by three different English words: "order", "type" and "class"; epistèmè is translated either by "science" or by "knowledge" by Grube and Reeve), and conversely, that the same English word can be used to translate different Greek words, again, not only from one translator to another (for instance, "understanding" is used by Jowett and Shorey to translate dianoia and by Grube and Reeve to translate noèsis, and also noûs), but by the same translator (as is the case for Grube and Reeve with the word "understanding" used to translate both noûs and dianoia). This is unavoidable since many words, especially abstract words, don't have a single well defined meaning but a whole range of meanings and that there is no reason for words in different languages to have in each case the exact same range of meanings as those which translate some of these meanings in other languages. The best example for that, at the heart of Plato's thought in his dialogues, is logos, the word describing what distinguishes human beings from all other animals, as can be seen by looking at its long list of meanings in the entry for this word in the LSJ.

In most cases, this is not a problem, but it becomes one with Plato, especially in texts such as the ones here examined, where, in full knowledge of what he is doing, he plays with these ambiguities of words, precisely to help us become conscious of the problem and understand that understanding is not in the words, but beyond words at the level of what he calls eidè and the relations between them expressed through logoi. The dilemma a translator is faced with in such a case is that one has to chose between always translating the same Greek word by the same English word (the choice explicitely made by Bloom), with the risk of using a word that doesn't cover some of the meanings Plato was playing with, and translating the same Greek word by different English words depending on the context, thus masking to the reader the fact that, in Greek, Plato was using the same word, often playing with its ambiguities. To avoid having to choose between these two equally unsatisfactory solutions in the case of Plato, there remains two not totally satisfactory options, which are not exclusive from one another: to leave the Greek word untranslated, which may be the least bad solution for such very sensitive words such as logos, eidos, idea or dialektikè, and/or to use notes to explain the problems and ambiguities of the underlying Greek word and mention for the reader other places where the same Greek word is used by Plato but translated by a different English word so one can compare the various uses of the Greek word with different meanings.

Two comments by some of the translators here mentionned point at some aspects of this problem. The first one is in Cornford introduction to his translation of the analogy of the line, about the problem posed by the word eikasia used to describe the state of mind (pathèma) associated with the lower segment of the line. It runs: "The lowest form of cognition is called eikasia. The word defies translation, being one of those current terms to which Plato gives a peculiar sense, to be inferred from the context. It is etymologically connected with eikon = image, likeness, and with eikos = likely, and it can mean either likeness (representation) or likening (comparison) or estimation of likelihood (conjecture). Perhaps 'imagining' is the least unsatisfactory rendering." What he says about this word should be extended to the three other word Plato uses for the three other pathèmata, even if they seem to pose less problems of translation (not of comprehensions, which is another matter altogether!) to translators: pistis, dianoia and noèsis, and helps understand why he waits till the end of the analogy to introduce the four words all at once: Plato indeed describes notions that are new but must do it with existing words, and as he doesn't want to create a "technical" vocabulary (see below) since he wants his readers to look for meaning beyond words, in the relations established between them, he chooses for each one of the four pathèmata ("faculties, affections, states (of mind), processes, conditions" depending on the translator; "that which befalls one, suffering, misfortune, emotion, condition, affection, feelings, troubles, symptoms, incidents, happenings, properties, accidents" according to the LSJ), which become moirai in the recalling of the analogy at book VII (see later in this note for the various meanings of moira), an existing word whose usual meaning(s) contributes something to the understanding of what he has in mind about the pathèma / moira he attributes it to, especially when compared to the words he attributes to the other ones, but which is not intended to become the "technical" name of it. And this is where the second comment comes into play, the note Grube introduces on the occurrence of noèsis at 534a2: "Before (i.e. at the end of Book Six, 511d-3) Plato called the highest mental process noêsis, which we translated as ‘‘understanding.” Now he calls it epistêmê (knowledge), and he applies the term noêsis to both the higher sections, that is, to the whole intelligible world, which we here translate as “intelligence.” The change is probably not careless but deliberate, and it illustrates his apparent dislike of a fixed technical vocabulary which he seems to have felt was at times an obstacle to understanding. At any rate he has just told us that we should not quarrel about words (533e)." Indeed, if Plato calls the highest segment noèsis in the analogy and epistèmè in the reminder, uses noèsis in the reminder to name something else, which includes but is not limited to what he calls by that name in the analogy, and talks of pathèmata in the analogy and of moirai in the reminder to refer to the same things, it is definitely to invite us to look for meaning beyond the words, and consider them as only shedding a partial light on what they name. There is indeed no possible doubts that what he calls epistèmè in the reminder is what he calls noèsis in the analogy since, in both cases, there are four pathèmata / moirai involved and the names of the three others are the same in both instances. And the fact that he uses noèsis to designate something else in the reminder is a way of suggesting that the word is fully adequate in neither case and should be understood from the whole picture of the analogy in its relations to the other words. And similarly, the change from pathèma, which suggests something that affects (paschein at the root of pathèma) the mind from the outside and is the result of something acting (prattein, the opposite of paschein) on it, a pragma ("fact / thing", derived from prattein in the same way as pathèma derives from paschein), to moira, meaning "part, portion, division", but also "share, fate, lot, destiny", and is the name of the goddess of fate, suggests that not everybody may access to all four "divisions / fates", an undertone that is lost in most traductions of the word which remain in a "geometrical" framework suggested by the analogy (in the quoted translations: division, part, section).
And to give another example of Plato plays with words to reach his goals, we may take another look at the four Greek words he uses in the analogy, from the mere standpoint of their "material / phonetic" structure. They are, from the lower to the higher, eikasia and pistis in the seen and dianoia and noèsis in the in the perceived by intelligence: as can be seen, the first name of each group of two ends with the suffix -ia which indicates a quality or an abstraction as such, to be taken for granted, while the second name of each group ends with the suffix -(s)is, wich indicates an activity. By these choices, Plato illustrates graphically / phonetically the ana ton auton logon ("according to the same ratio(nale)" which should preside over the segmentation of the two segments resulting from the first segmentation: the first segment in each case indicate a state of mind in which one takes at face value what one sees with the eyes (segment of the seen) or grasps by the mind (noûs) (segment of the perceived by intelligence) without realizing that it is only "images" (eikones) of what acts upon the eyes (images on the retina) or mind (words as "labels" arbitrarily attached to what they purport to desigante: in the Cratylus, there are 19 occurrences of the word eikôn in references to words / names as "images" of something else), while the second segment implies an active participation of the one subjected to such a pathèma ("state of mind") in order to intrepret what one knows to be a mere "image" with the help of dialogue with other people. The choice of noèsis in the analogy was probably influenced by this wish of Plato to illustrate in the words themselves the instructions given by Socrates for the second segmentation, even though he knew it was not totally satisfactory. But it was not wrong either since the reminder in book VII uses the word for more that what it desigantes in the analogy. It is up to us to understand what he has in mind for the pathèma associated with the second segment of the perceived by intelligence, not from the word noèsis alone, but from the fact that it is only a part of what is called noèsis in a broader sense in the reminder, a part wich may also be called more specifically epistèmè ("knowledge") and which excludes that part of noèsis which is called in both cases dianoia ("discursive / wandering thought"), that is, from the relations he evidences or suggests between these words and other words used in the analogy. This example shows why it is almost impossible to render such subtle use of words in any other language.

Let us now come to the word that poses the most problems for translators, which is the word that is precisely central to this entire passage, and more generally to the understanding of Plato: it is the word eidos (the word idea does not appear in these lines), as shown by the list below of the 9 English words or groups of words (17 in French) which, in the analogy of the line alone, where it appears 7 times, have been used to translate it. In this list, the words used by the translators listed in the table to translate eidos are presented in alphabetical order, distinguishing, for "form" and "idea", the cases in which the word is used without a capital letter and the cases in which it is used with a capital letter when it is not the first word in the sentence, a way for some translators to emphasize that the word is then taken in a specifically "Platonic" sense (according to their understanding) referring to to a supposed "theory of Forms/Ideas" that they attribute to Plato:

eidos = class, figure, form, Form, idea, kind, kind of things, order of things, type
(and in French, in the translations used in the French version of this page: aspect, classe, espèce, espèce d'êtres, façon de penser, figure, Forme, forme, forme idéale, genre, Idée, idée, idée pure, image, ligne figurée, monde, nature essentielle, ordre, sorte)

Jowett uses 3 different words (idea, kind, form) to translate it, Shorey 5 (type, idea, kind, form, class), Cornford 5 (order of things, Form, kind, figure, class), Grube 4 (Form, kind, figure, class), Reeve revision of Grube translation 3 (kind (of things), form, figure), Reeve 2 (form, kind). Alone, Bloom, by deliberate choice for his whole translation to always translate the same Greek word by the same English word, uses the same word, "form", for all 7 occurrences of eidos.
The number of different English words or groups of words used by the same translator to translate these 7 occurrences in 3 pages of the same Greek word is most often 4 and varies between 3 and 5. How, under these conditions, can a reader who does not have access to Greek know that, behind these different English words, there is the same Greek word and that, even if it is not taken by Plato each time in exactly the same sense, it seemed important to him to draw in each case, by the use of this same word, resonances that disappear in the translation?
The problem is that translators persist in thinking that eidos has several meanings, which is true, well compartmentalized from one another (as for instance in English "crane", which can designate either a bird or a machine for hoisting and moving heavy objects by means of cables attached to a movable boom), which is false, and that this word sometimes has a "usual" meaning in Plato ("class / figure / kind / type... "), sometimes a "technical" meaning ("Form / Idea", with or without the first letter capitalizes, in the supposedly "Platonic" sense) impervious to the usual meaning and that it is more important to emphasize the differences in the translation than to preserve the possible ambiguities, most probably intended by Plato to encourage the reader to ask oneself questions and thus to progress in the understanding of what he seeks to make understood,  and not of what his translator tries to make him say. In doing so, they forbid themselves to understand Plato, who, although very attentive and rigorous in the use of words, was, as I said earlier after quoting Grube saying the same thing, never adept at a fixed technical vocabulary, preferring to play with the polysemy of words rather than seek (in vain, because it is an impossible task) to rule words in order to bend them to his needs (as Aristotle tried to do). Now what Plato tries to make us understand by persisting in using words related to sight (such as eidos and idea, both formed on roots meaning "to see" and of similar meaning) to speak of the intelligible realm that he himself baptizes noèton, and in which what he calls dianoia is practiced, whereas he had at his disposal words such as noèma (which is found for instance at Symposium, 197e5 and more to the point 5 times in the discussion between Socrates and Parmenides at Parmenides, 132b-c, where Socrates substitutes it for the word eidos used by Parmenides) or dianoema (which is found, for instance, at Protagoras, 348d3, at Symposium, 210d5 and at Republic, VI, 496a6), it is precisely that, far from there being watertight partitions between the two meanings of this word (usual and "Platonic"), there is continuity from one to the other, from the individual visible appearance (primary meaning) to the community of visual appearance that makes it possible to identify (and therefore to name) a "species / genus / kind" of things to the community of criteria of resemblance of any kind, visible or non-visible, which characterize sets of "things", material or immaterial,  to which the same name is attributed. Those who want eidos, et idea, in the supposedly "Platonic" sense, which they consider as synonyms in this sense, to represent what would constitute the ultimate immaterial reality of all things, do not realize that to accept this is to suppose that Plato would have chosen, in order to designate in the intelligible realm what has the most "reality", what "is" in the fullest sense of the term (according to their "theory") two words which, in the visible realm, designate precisely what has the least, a simple "appearance" for the eye! To present new ideas in a field that was particularly difficult because of its abstraction, in which he could see that he had the greatest difficulty in making himself understood, even by his most brilliant disciple, Aristotle, the field of what we call "ontology" today, such a choice would not only have been disastrous, but also stupid! It is not by completely reversing the meaning of a word in order to move from its proper meaning to an analogical meaning, here from a meaning in the register of sight to an analogical meaning in the register of the intelligible, that one increases one's chances of being understood! That others believed that this was what he had done is their problem, not his. What he seeks to make us understand is, on the contrary, that in both registers, that of the intelligible as well as that of the visible, the limits of our human nature, those of our eyes in the visible, those of our mind (noûs) in the intelligible (noèton), do not allow us to apprehend what acts on them as it is, but only according to the "appearances" (eidè in the primary sense) that it takes for human vision in one case, for human intelligence in the other. And I am not refering to the specific limitations of a particular individual, who may be nearsighted, colorblind or have this or that vision defect, or heavy-minded, or even mentally handicapped, as far as intelligence is concerned, but limitations that are inherent in human nature, such as the fact that the eyes are only activated by a certain range of electromagnetic wave frequencies and perceive only the light reflected in that range by material surfaces, not sounds or smells, and that human intelligence can only express itself through words related to one another in vocal, graphic or inner logoi, which can only give us access to relations between "beings", not to the "*** themselves" (ta auta) that are at the origin of the activity of our senses and mind. In short, whether with views through the eyes or "views" of the mind, what we have access to are always "appearances" (eidè), which are presented as colored images for the eyes, words taking meaning in logoi establishing relations between them for the intelligence. The key to understanding what Plato means by eidos and the difference he makes between eidos and idea is given by a passage at the beginning of Book X of the Republic, the discussion on the different kinds of beds, where "politically correct" Platonists are reluctant to look for it, so much does this discussion on trivial material creations of human craftsmanship (dealing with tables and beds) seem to them far removed from the idea they have of Plato's eidè / ideai, which they prefer to seek on the side of the beautiful, the just, the good, etc., whereas it is precisely to put himself within their reach that Plato sought his examples in the products of human craftsmanship for which the purpose that makes them intelligible is obvious to all, more so in any case than what makes men (anthrôpoi) intelligible to themselves or what makes justice intelligible to them, which is nothing other than the idea(l) of the incarnate man: a table is intended for putting things on and a bed is intended for people to lie on it to rest or sleep, this is the "idea" (idea) that inspires their makers, what they "look towards" (with the "eyes" of the mind) when they build such a piece of furniture (cf. Republic X, 596b6-10). As for the eidè, their "definition" is provided at the beginning of this discussion on tables and beds: "we are in the habit of positing for our own use some eidos, unique in each case, for each of the pluralities upon which we impose the same name" (Republic X, 596a6-7). An eidos is, for a given person, what a word he or she uses refers to: it begins with visual images when we are children and learn to speak (the prisoners in the cave who give names to shadows, cf. Republic VII, 515b4-5) and it evolves over the course of our lives by gradually replacing purely visual criteria with criteria in the intelligible realm. These eidè are therefore different from one person to another, and for the same person from one moment of one's life to another, but they all have the same targets, the ideai (the stars in heaven outside the cave), which are independent of what each person apprehends through eidè at a given time in one's earthly life, but which are not yet the "*** themselves" (ta auta), inaccessible to human intelligence in what they are "in themselves", and which are still only "appearances" for human intelligence at its best, therefore conditioned only by its limits, which guarantees their objectivity. And, for the record, it was the choice I made for version 2 of my translation (in French) of the analogy of the line (put online in October 2012) to translate eidos everywhere by the same French word, "apparence" (its primary meaning), that opened the way for me to understand this analogy, and more generally Plato, in particular by the connection that this choice made possible between the two successive uses of eidos within a few lines, at 510d5 about horômenois eidesi ("visible eidè ("appearances")) and at 511a3 about noeton eidos ("intelligible eidos ("appearance") "), whereas, as can be seen in the table above, all the translators I consulted except Bloom translate these two successive occurrences of the same Greek word eidos by two different words! Eventually, Plato did not seek to give eidos et idea a new, specifically Platonic meaning, the same for both words, but to push the analysis of the evolution of their meanings from an individual meaning exclusively related to sight ("appearance (for sight)") to a collective meaning of "genus, kind, species..." where the criteria for grouping can be of any kind, not only visual, seeking to show how this could make it possible to understand the way logos (language) works and to explain the reasons why we manage to understand one another despite the perfectly arbitrary nature of the choices of words at the outset, by distinguishing in this process a subjective element, therefore different from one person to another, the eidè, and an objective element, the same for all, but which remains an inaccessible target, the ideai, the choice to specialize each of these two words with very similar meanings and both built on roots referring to sight (because this is where everything takes its source for human beings who begin their lives chained at the bottom of the cave and having no other apprehension of the world around them than that which sight gives them, but are already capable of speaking and dialoguing), being largely arbitrary since what counts is not the pre-existing meaning(s) of each of these two words, but the difference between them that this choice induces.
And what we must not lose sight of in translating Plato, especially in a passage such as this one, is that he tries to bring his readers to a new understanding of their ways of apprehending the world that surrounds and affects them, and that, in order to do this, he cannot therefore find in the language he speaks words that correspond exactly to what he has in his mind, to the eidè that he associates with these words and to the ideai that he seeks to reach by using them. As we saw earlier in this note, this is particularly glaring here in the case of the four pathèmata ("affections / states of mind") that he associates with the four sub-segments of the line. If he gives them all at once and only at the end of the analogy, it is in the hope that what has been said before will enable us to understand them in the sense he gives them, and not in order that we may rectify our previous understanding by relying on the usual meaning which they hitherto had. Of course, if he preferred to use pre-existing words rather than create neologisms formed for the occasion (as he does in other places), it is so that these usual meanings contribute to shedding light on these new notions that he introduces, but on condition that these previous meanings do not obscure the meaning he seeks to give them, but contribute only to clarify and specify it. Of course, the exercise is difficult, and we will never be sure that we are reaching the eidè that Plato had in mind by using these words in this context, but what interests Plato and what he expects from his readers is precisely this work of seeking understanding, the target of which is not the eidè that he had in mind, but the ideai he was aiming at. And this work is made even more difficult for us by the fact that his language is not ours and that we seek to understand it with the help of dictionaries that take into account a state of the language to be translated that is later than that which it had in his time and influenced not only by his own writings, but by those of other later authors, including commentators on his writings, and that we can never be sure that we have found the meaning(s) of the words he uses that were current before he wrote and whose range of meanings he enriched. Thus, to give the names of the four pathèmata all at once at the end is to encourage us to try to understand them in their relations to one another in the light of the explanations given before, and not to try to understand them individually independently of one another with the help of dictionaries. As shown by the experiment with Meno's slave, where Socrates gives his interlocutor the name of the line ("diagonal") whose length constitutes the answer to the question posed (find the length of the side of a square with an area twice that of a given square) only at the end of the experiment, once the slave has shown on the figure drawn by Socrates without naming it the line that he understood to be the answer to the question, it is not the name that makes us know the thing to which it is attributed, but it is once we know the thing that we feel the need to give it a name to facilitate the conversation.