(1.1)

One textual difficulty is that the notion of ‘change of meaning’ is inconsistent with the context of the stasis-passage.Footnote ⁹ When Dionysius says ‘as they changed the words usually said of things, people thought it fit to call [things] in a different way’, he seems to imply that people changed the words they used to name things. Let us name this linguistic phenomenon ‘word change’: Given an item x, at a time T¹ a word N¹ means x; and at a time T², N² (and not N¹) means x. In Dionysius’ reading, we have an item x and two names, N¹ and N². Item x remains invariable. However, the name that names x shifts. First, at a time T¹ people customarily name x N¹ and then at a time T² people abnormally name x N².

When we think of a word changing its meaning, however, we think of semantic drift. ‘Nice’, for instance, is today a commendatory adjective and means agreeable. But in the late thirteenth century, it was pejorative and meant silly. In contrast, ‘silly’ was in the thirteen century a commendatory adjective and meant blessed, while today it is a pejorative adjective and means lack of judgement. Semantic drift occurs when the senses of a word vary, disappear or new senses crop up, transforming its overall meaning. In the case of ‘silly’, the view that the blessed are inoffensive and naïve led to the assumption that the blessed are helpless and therefore lack judgement. Whereas in the case of ‘nice’, the view that silly people are harmless led to the assumption that the silly are inoffensive and therefore agreeable.Footnote ¹⁰

In a semantic drift, a word remains the same, but it switches an old meaning for a new one. In this case, a word—the phonetic or written mark—is a subject of change, and the change consists in the substitution of an old meaning for a new one. Accordingly, when we describe a semantic drift, the subject of description is a word, and the predicate contains (at least) two different meanings with different time indexations. For instance, ‘nice’ today means agreeable, but in the late thirteenth century it meant silly. This description is about the word ‘nice’, an entity which we can characterize as having four letters. For this reason, the description does not use, but mentions, the word ‘nice’ and places it—following contemporary academic usage—between single quotation marks.Footnote ¹¹ By contrast, the description attributes meanings to the word ‘nice’. These meanings—agreeable and silly—are not entities we can characterize as having nine or five letters. Accordingly, the description uses the words ‘agreeable’ and ‘silly’ and does not place them between single quotation marks.

Wilson suggests that the stasis-passage cannot be about semantic drift because semantic drift takes a long time, and the passage is about a sudden change.Footnote ¹² However, semantic drift need not take centuries to develop; the most interesting cases, indeed, are sudden and traumatic. In Tsarist Russia, for instance, the word ‘kulak’ meant a rich peasant. But after the October Revolution it came to mean a peasant who opposed collectivization. Similarly, the word ‘proletarian’, which used to mean a working-class person, came to mean a supporter of the government or Communist Party. In the Soviet Union, it was good to be a proletarian, as it was also to be a communist. But not all communists were proletarians; consequently, ‘proletarian’ drifted to mean a supporter of the government, and party-leaders such as Lenin (who was of noble origin) were able to be both a communist and a proletarian.Footnote ¹³

Semantic drift is a social phenomenon of linguistic behaviour that may respond to a semantic-linguistic incentive, and sometimes also to socio-cultural pressure. This phenomenon consists in a word—the phonetic or written mark—remaining invariable but switching meanings. For instance, while Marx intended the word ‘proletarian’ to mean working-class person, the Soviets came to use the word as meaning supporter of the government and the communist party. We may also describe semantic drift as a word remaining invariable but switching the conventions that govern its use. A working description of semantic drift may be the following: Given a name N, at a time T¹ N means x, and at a time T² N means y. Semantic drift differs from word change, the linguistic phenomenon Dionysius of Halicarnassus attributes to the stasis-passage. In semantic drift, words—the phonetic or written marks—do not change, but their meaning does; whereas in word change, words—not their meaning—change.

Despite these differences, word change and semantic drift are related phenomena. Semantic drift may follow word change and vice versa. Take ‘nice’ and ‘silly’ as an example. People used to call the agreeable (that is, those who easily agree) ‘nice’ and the brilliant ‘silly’. However, they began to call the agreeable ‘silly’ and the brilliant ‘nice’. We can equally describe these semantic changes as word changes or as semantic drifts. We can say that people used to call the agreeable ‘nice’, but now they call them ‘silly’; and correspondingly they used to call the brilliant ‘silly’, but now they call them ‘nice’. These are word changes. But we can also say that ‘nice’ meant first agreeable and later brilliant, while ‘silly’ meant first brilliant and later agreeable.Footnote ¹⁴ These are semantic drifts. What does Thucydides have in mind in the stasis-passage, semantic drift or word change as Dionysius believed? To answer this question, we must look at Thucydides’ examples.

(1.2)

When Thucydides explains the stasis-passage by means of examples (all of which are cited in Appendix II), we do encounter social phenomena, though nothing resembling a description of word change, semantic drift or linguistic behaviour. His first example (Ex.1) is:

This example describes a social phenomenon, with the passive verb ἐνομίσθη suggesting a general agent, ‘people’. However, Thucydides does not intend the subject of this description, τόλμα ἀλόγιστος, to be a linguistic expression and to have fourteen letters. He does not mention, but uses, the expression “τόλμα ἀλόγιστος”, and τόλμα ἀλόγιστος is an action, the sort of action we deem good or bad, a moral action.

By contrast, ἀνδρεία φιλέταιρος is here intended neither as a linguistic expression nor as the referent of one. To be precise, ἀνδρεία φιλέταιρος is what people, in unusual circumstances, take τόλμα ἀλόγιστος to be. While in this text τόλμα ἀλόγιστος is an action type, ἀνδρεία φιλέταιρος is here a belief about τόλμα ἀλόγιστος. And this belief, as both ἀνδρεία φιλέταιρος and the passive verbal form ἐνομίσθη (be believed/esteemed) suggest, involves a moral assessment.Footnote ¹⁵ Clearly, Ex.1 is about moral assessment, not about linguistic behaviour; it is about a drift in moral assessment, not a semantic drift.

The question arises whether Ex.1 is a representative case of what the stasis-passage says. It must be, since Ex.1 is the first of ten consecutive examples that share the same verbal form, ἐνομίσθη. This verbal form is explicit in Ex.1 but implicit in the eight following examples, all of which are consistently about types of moral actions and excellence (or wickedness) in performing moral actions.Footnote ¹⁶ This consistency indicates that the stasis-passage must be about drift in moral assessment.

Thucydides provides even more examples: at least fifteen distributed through 3.82.4–3.83.4, all of which involve types of moral actions and excellence (or wickedness) in performing moral actions.Footnote ¹⁷ Eight display unequivocal verbal expressions of moral assessment; for instance ‘being prized’ (ἐπῃνεῖτο), ‘accept … not out of noble character’ (ἐνεδέχοντο … οὐ γενναιότητι), ‘being called righteous’ (δεξιοὶ κέκληνται), etc.Footnote ¹⁸ Some of these expressions do exhibit vocabulary that may suggest talk about meaning and semantics, as, for instance, κέκληνται. However, καλέω can have evaluative and moral undertones. To call people righteous (δεξιοί) or honourable (ἀγαθοί) is to praise them and to pass judgement on them; for example οἱ πολλοὶ κακοῦργοι ὄντες δεξιοὶ κέκληνται ἢ ἀμαθεῖς ἀγαθοί, 3.82.7 (Ex.17). The other seven examples claim that moral actions and abilities that prevail in stasis do not prevail in normal circumstances, and this claim presupposes a moral assessment.Footnote ¹⁹

Thucydides’ multiple examples are consistent with drift in moral assessment, not with semantic drift. Since these examples intend to explain the stasis-passage, Thucydides must take this line to be about drift in moral assessment. And if that is the case, then the traditional translations and readings that take the stasis-passage to be about meaning must be wrong, saddling Thucydides with the inconsistency of explaining cases of semantic drift by offering examples of drift in moral assessment. The stasis-passage, therefore, cannot be about semantic drift and must be about drift in moral assessment.

(1.3)

Where exactly does the traditional reading of the passage go wrong? LSJ lists ‘meaning’ as a definition of ἀξίωσις, but the only example it provides for such definition is Thuc. 3.82.4.Footnote ²⁰ As it appears, the traditional reading—for example Jowett's, Crawley's and Foster's translations—defies the etymology of ἀξίωσις, which speaks for valour or moral value.Footnote ²¹ The action nominal suffix -σις suggests that ἀξίωσις means valuation. Valuation, unlike meaning, is a meaning of ἀξίωσις compatible with Thucydides’ examples. However, the traditional reading seems to have been so entrenched by Hobbes's (T4) time that he respects the etymology of ἀξίωσις but inserts the term ‘signification’. He translates: ‘The received value of names imposed for signification of things was changed into arbitrary’.Footnote ²²

Now, if ἀξίωσις in 3.82.4 means value or valuation, we need to modify other features of the traditional translation of the stasis-passage. Jowett's (T1) and Foster's (T3) translations, although inconsistent with the subsequent examples, are internally consistent. Consistently with their rendering of ἀξίωσις as meaning, Jowett and Foster read τῶν ὀνομάτων and τὰ ἔργα as a dichotomy between words and the things they mean. However, if ἀξίωσις means value or valuation, τῶν ὀνομάτων–τὰ ἔργα must be a dichotomy between words and actions. And actions are the proper object of moral value or valuation. Connor's translation (T5) gets this right: ‘And they modify at their discretion the customary valences of names for actions’;Footnote ²³ similarly Mynott's (T6): ‘Men assumed the right to reverse the usual values in the application of words to actions’.Footnote ²⁴

The stasis-passage is about value or valuation, and Thucydides explains this passage by appealing to examples of drift in moral assessment. Unfortunately, the role words play in this picture is unclear. One possibility is that drift in moral assessment may trigger semantic drift. None the less, Thucydides might well have something else in mind.

Before discussing the role of the words in the stasis-passage, we need to work out what precisely the passage says: whether ἀξίωσιν in this passage means value or valuation, whether ἐς τὰ ἔργα complements ἀξίωσιν or δικαιώσει, and what δικαιώσει and ἀντήλλαξαν mean. I clarify all these issues in the next section. But to do all this, we need first to study some of the examples that follow the stasis-passage.

(2.1)

Let us take a look at the stasis-passage (henceforth, S-P) and Ex.1 together:

The preceding context dwells on how human nature responds to the different circumstances of peace and stasis (3.82.1). In times of peace, people lack the excuses, motives, constraints and necessities they have in times of stasis (3.82.2; 3.82.3). For this reason, people have better judgement in peace than in stasis. We have here three different interconnected antitheses: peace—stasis, prosperity—necessity, sound judgement—bad judgement. The first example assumes these antitheses:

This expanded reading of the first example illuminates part of the vocabulary of the stasis-passage, εἰωθυῖαν and ἀντήλλαξαν. The standard translation of εἰωθυῖαν is ‘usual’, ‘accustomed’, or ‘ordinary’. However, the context between brackets suggests that what the text calls ‘usual’ coincides with peace, prosperity and good judgement. Consequently, εἰωθυῖαν in this text is not any old usual, but what good judgement grasps and what we expect to be normal. This is a ‘normal’ which contrasts with judging things wrongly, the abnormality of stasis and the undesirability of necessity. More than usual, “εἰωθυῖαν” in the stasis-passage means normal.

Normal (εἰωθυῖαν) in the stasis-passage contrasts with an implicit ‘abnormal’. This is key to understanding ἀντήλλαξαν. The verb ἀνταλλάσσω demands two complements, one in the accusative case, the other in the dative (τί τινι). Here the required accusative is τὴν εἰωθυῖαν ἀξίωσιν; however, τῇ δικαιώσει does not appear to be the required dative. It lacks the expected qualification ‘abnormal’, and fails to be the antithesis of τὴν εἰωθυῖαν ἀξίωσιν. Hence there appears to be an ellipsis in the stasis-passage. The text does not explicitly express, but assumes, a dative expression antithetical to τὴν εἰωθυῖαν ἀξίωσιν, something like τῇ οὐκ εἰωθυίᾳ or τῇ ἀήθει ἀξιώσει.

Thucydides speaks of exchanging or substituting (ἀντήλλαξαν) the normal ἀξίωσιν—which results from good judgement, prosperity and peace—for an abnormal ἀξιώσει—which results from lack of judgement, necessity and stasis. As Wilson observes, translations that render ἀντήλλαξαν as ‘to change’ overlook ἀντ-. A change is merely a process in which an item becomes different in some respect or another. By contrast, an exchange is the substitution of an item for an altogether different item.Footnote ²⁵ In this case, we have the substitution of a certain sort of ἀξίωσις or way of regarding things (ἐνομίσθη) for an altogether different sort of ἀξίωσις or way of regarding things (ἐνομίσθη).

The question remains: if τῇ δικαιώσει is not the dative complement of ἀντήλλαξαν, what is it? It must be some sort of adverbial modifier or dative of circumstances. However, it is unclear which dative or sort of adverbial modifier. Interpreters and translators have proposed readings such as ‘at their will and pleasure’,Footnote ²⁶ ‘at their discretion’,Footnote ²⁷ ‘at their pleasure’,Footnote ²⁸ ‘arbitrary’Footnote ²⁹ and ‘by the arbitrary construction’.Footnote ³⁰ This family of translations probably developed from the fact that δικαίωσις sometimes means the action of claiming a right, not to mention that the etymology of the word harks back to opinion. However, it is a stretch to move from ‘claiming what is right for oneself’ to ‘personal whim’ and ‘random (choice)’.

A different and more etymologically minded family of translations stems from Dionysius of Halicarnassus. He paraphrases ‘they considered it right to call things differently and transposed the accustomed words said for the actions’ (τά τε εἰωθότα ὀνόματα ἐπὶ τοῖς πράγμασι λέγεσθαι μετατιθέντες ἄλλως ἠξίουν αὐτὰ καλεῖν, Thuc. 29.30–1). Translations of this family read δικαιώσει in the stasis-passage as follows: ‘claiming a right’, ‘as one judges right’, ‘men claimed the right to’,Footnote ³¹ ‘as they thought proper’,Footnote ³² ‘as men thought fit’,Footnote ³³ ‘men assumed the right to’.Footnote ³⁴

These two families of translations are objectionable. The first—‘arbitrary’—is far too distant from the etymology of the term. Moreover, the only example of such meaning of δικαίωσις LSJ quotes is the passage we are discussing at Thuc. 3.82.4. Just like Dionysius’ reading of ἀξίωσις, this is an ad hoc reading of δικαίωσις. Further, these translations wrongly construct an adverbial modifier, or dative of circumstances, which should qualify the verb and action at issue—namely, ‘substituted’ (ἀντήλλαξαν). This mistake also applies to the second family of translations. By translating ‘as one judges right’ or ‘as they thought right (proper, fit)’, this second family of translations applies the dative of circumstances ‘judging (or claiming) what is right’ to the subject, not to the verb ‘substituted’ (ἀντήλλαξαν), as one may expect.

The context suggests a better translation of τῇ δικαιώσει. In Ex.1, ἐνομίσθη manifestly shows that the stasis-passage involves moral assessment. If that is the case, then τῇ δικαιώσει in that passage should mean ‘judging what is right’, ‘in passing judgement’, as Wilson has suggested.Footnote ³⁵ ‘In passing judgement’ adverbially and circumstantially qualifies ‘substituted’ (ἀντήλλαξαν). Thucydides is therefore saying ‘people substituted when passing judgement’ or ‘people substituted in passing judgement’.

One mistake leads to another. Some translators who did not construe τῇ δικαιώσει with ἀντήλλαξαν have construed ἐς τὰ ἔργα with τῇ δικαιώσει. Gomme, for instance, translated: ‘the customary meanings of words were changed as men claimed the right to use them as they would to suit their actions: an unreasoning daring was called courage’ (my emphasis) (T10).Footnote ³⁶ Commentators have complained about this construal, insisting that ἐς τὰ ἔργα goes with ἀξίωσιν.Footnote ³⁷ This point is crucial. If ἐς τὰ ἔργα expresses the goal of ἀξίωσιν, then clearly ἀξίωσιν expresses action, and it means valuation, attribution or assessment of value. Pace Hogan and Wilson, ἀξίωσιν means not value but valuation.Footnote ³⁸ We can, at this point, provisionally conclude that the stasis-passage claims:

We still need to explain what τῶν ὀνομάτων in these lines means and how it qualifies ‘valuation’ (ἀξίωσιν).

(2.2)

τῶν ὀνομάτων is an attributive genitive that limits the meaning of ἀξίωσιν. What sort of attributive genitive is it? What does ὀνομάτων mean? Thucydides is certainly talking about words without specifying any category, such as nouns and verbs. However, the question about the meaning of ὀνομάτων is relevant because it is possible to read ‘words’ in different ways. There are two issues here. First, when Thucydides mentions τῶν ὀνομάτων, it is unclear whether he means single words, phrases, sentences, or all of these. Second, it is also unclear whether he means meaningful linguistic expressions or only phonetic/written marks detached from their semantic content, meaning, or referent.

Let us begin with the first issue. We can describe Ex.17 as applying single words to people (δεξιοί, ἀγαθοί) to praise or morally assess them. Most examples, however, talk about applying phrases to actions or people to praise, blame, or morally assess them. But the point is always to show ways in which people pass moral judgement. And when people pass moral judgement, they not only produce words and phrases but also produce or imply declarative sentences. Take Ex.1 and Ex.17. According to Ex.1, people in stasis consider reckless audacity to be courage in the protection of partisan interests. This means that people in stasis must at least be committed to claiming that ‘such and such acts are instances of courage in the protection of partisan interests’. The same applies to Ex.17. According to Ex.17, people in stasis call evil-doers righteous. All this implies that people in stasis must at the very least be committed to claiming that ‘such and such an individual is righteous’. If this is the case, then τῶν ὀνομάτων in the stasis-passage must refer to all sorts of linguistic qualifications people attribute to actions when they express moral assessment, including words, sentences, statements and descriptions.

Similar considerations apply to the second issue. As we saw, the expression τῶν ὀνομάτων refers to the linguistic qualifications, descriptions and sentences people use to express moral assessment. But if this is the case, then τῶν ὀνομάτων must refer to meaningful words, phrases and sentences, not to phonetic/written marks without their semantic content.

This treatment of linguistic expressions confirms that Thucydides is concerned neither with word change nor with semantic drift. Word change is about a particular referent (or meaning) x, to which a phonetic/written mark¹ refers (or means) at a time T¹ and to which a different phonetic/written mark² refers (or means) at a time T². Semantic drift is about a particular phonetic/written mark, which refers to (or means) x at a time T¹ and refers to (or means) y at a time T². To talk about these linguistic phenomena, one must analyse a word into a phonetic/written mark and its referent (or meaning). But since Thucydides is considering words as meaningful linguistic expressions, he is not analysing them into phonetic/written marks and their semantic content. In the text, there is no indication that Thucydides is analysing words. And it seems we can make sense of the text without importing such analysis. For these reasons, Thucydides does not appear to be talking about word change and/or semantic drift.

Now that we know that ἀξίωσιν in the stasis-passage means valuation and that ὀνομάτων refers to meaningful words (terms, descriptions, sentences), we can finally answer the second question. What sort of attributive genitive is ὀνομάτων? In other words, how does ‘words’ qualify valuation? For sure, ὀνομάτων cannot be an attributive genitive of material, content, praise, or value, let alone a partitive genitive. What is it then?

Let us assume for the sake of argument that ἀξίωσιν means value. In this case, τῶν ὀνομάτων should be (1) a genitive of possession or (2) an objective genitive, and ἀξίωσιν τῶν ὀνομάτων would mean ‘the value words have’. On the other hand, if ἀξίωσιν means—as I think it does—the activity of valuation, τῶν ὀνομάτων should be (2) an objective or (3) a subjective genitive. In cases (1) and (2), words would be the objects that suffer the activity of valuation. But again actions, not words, are the objects of valuation in this text. In case (3), words would be the subjects that do the activity of valuation. But humans, not words, are agents of valuation.Footnote ³⁹

We can confirm that τῶν ὀνομάτων is neither a subjective genitive nor an objective genitive—it is not a genitive of special determination—nor a genitive of possession. Consequently, we cannot translate ἀξίωσιν τῶν ὀνομάτων as ‘valuation of words’, because ‘of’ suggests that words are the object or possessors of valuation. What sort of genitive is it? τῶν ὀνομάτων certainly determines ἀξίωσιν and it could be a genitive of general determination, such as τὰς ξυμφορὰς … τῶν βουλευμάτων (Soph. OT 44–5), ‘the events derived from the resolutions’.Footnote ⁴⁰

The genitive of general determination provides any possible determination derived from the context or meaning of the words involved.Footnote ⁴¹ This suggests that any interpreter has leeway to construe τῶν ὀνομάτων as their interpretation requires. However, in my view Dionysius was right to think that the passage aspires to high literary airs and that Thucydides forced the grammar. Following Dionysius, I submit that τῶν ὀνομάτων is a genitive of apposition, also called of inherence, epexegetic, or of explanation, a poetic construction also frequent in prose.

The boundaries between the genitive of general determination and the genitive of apposition are blurry. However, the different uses of the genitive of apposition help clarify the construction of ἀξίωσιν τῶν ὀνομάτων. Usually the head-term and the genitive-modifier in apposition refer to the same item, though sometimes the genitive-modifier refers to something inherent to the referent of the head-term. This ambivalence explains the blurry boundary between these genitives. ἀξίωσιν τῶν ὀνομάτων is perhaps a blurry case. The genitive of apposition specifies an individual item that falls under the extension of the term it qualifies, as in Ἰλίου … πόλιν (Hom. Il. 5.642).Footnote ⁴² But sometimes the genitive-modifier gives new meaning to the term it qualifies, as in ἕρκος ὀδόντων, which means ‘the teeth as a fence’,Footnote ⁴³ because the inherent feature ὀδόντων becomes a salient feature and more prominent than ἕρκος.Footnote ⁴⁴ Sometimes the head-term and the genitive-modifier are synonyms of sorts, as in θανάτοιο τέλος, ‘death as the ultimate end’ (Hom. Il. 3.309, 9.411, 13.602).Footnote ⁴⁵ And sometimes the genitive of apposition narrows the meaning of the head-term by referring to a property of the head-term's referent, as in βίη Διομήδος, ‘mighty Diomedes’ (Hom. Il 5.781).Footnote ⁴⁶ However, the emphasis on the referent of the genitive-modifier disappears in prose,Footnote ⁴⁷ as in πηγῆς ὄνομα, ‘the noun “fountain”’ (Pl. Cra. 402c6–7), where πηγῆς tells us that the ὄνομα in question is the noun “πηγῆς”.

Our case, ἀξίωσιν τῶν ὀνομάτων, is akin to πηγῆς ὄνομα. But it is even more proximate to cases where a common word in the genitive narrows the meaning of a general or less common word. In this usage, the genitive-modifier defines the application of the head-term: for example καὶ τοι πῶς οὐκ ἀμαθία ἐστὶν αὕτε ἡ ἐπονείδιστος, ἡ τοῦ οἴεσθαι εἰδέναι ἃ οὐκ οἶδεν;, ‘however, how is this not the most reprehensible ignorance, the one of believing that one knows what one does not know?’ (Pl. Ap. 29b1–2).Footnote ⁴⁸ Here the genitive τοῦ οἴεσθαι εἰδέναι specifies one among many kinds of ignorance. Similarly, τῶν ὀνομάτων in the stasis-passage seems to specify one among different possible sorts of valuation.

There is, however, a difference between Ap. 29b1–2 and the stasis-passage. τοῦ οἴεσθαι εἰδέναι indicates what the ignorance in question essentially consists in. By contrast, τῶν ὀνομάτων indicates an inherent, non-essential feature of the valuation at stake. In the stasis-passage, Thucydides refers to valuations that are distinctive because people express them in words, that is, single words, phrases and sentences. We find better examples in Thucydides: for instance ἡ δὲ διαγνώμη αὕτη τῆς ἐκκλησίας τοῦ τὰς σπονδὰς λελύσθαι, ‘the assembly's decision to break the truce’ (1.87.6), where the genitive “τοῦ τὰς σπονδὰς λελύσθαι” elaborates on διαγνώμη by explaining the content of the decision. An even better example is Thuc. 7.42.2:

τοῦ ἀπαλλαγῆναι cannot be an objective genitive: the text is not about ‘an end to the removal of the danger’. Ending such removal may involve the continuity or restitution of the danger. Rather τοῦ ἀπαλλαγῆναι narrows the meaning of πέρας by introducing an inherent feature which distinguishes that πέρας from any other πέρας. Thucydides, therefore, speaks about the resolution of a situation different from any other resolution. And this resolution is different because it involves or coincides with the removal of a danger.Footnote ⁵⁰

Just as τοῦ ἀπαλλαγῆναι narrows the meaning of πέρας μηδέν in 7.42.2, τῶν ὀνομάτων narrows the meaning of ἀξίωσιν in the stasis-passage. The genitive τῶν ὀνομάτων introduces an inherent feature which distinguishes one ἀξίωσιν from any other ἀξίωσιν. The valuation of the stasis-passage is thus a valuation that has words as an inherent distinctive feature. How should we then translate ἀξίωσιν τῶν ὀνομάτων? I translated πέρας μηδὲν … τοῦ ἀπαλλαγῆναι τοῦ κινδύνου as ‘no resolution until/before the removal of the danger’ to disclose the relationship between πέρας μηδὲν and τοῦ ἀπαλλαγῆναι. I propose to do the same with ἀξίωσιν τῶν ὀνομάτων and suggest ‘the normal valuation (τὴν εἰωθυῖαν ἀξίωσιν) [expressed/made with] words (τῶν ὀνομάτων)’ or just ‘the normal valuation (τὴν εἰωθυῖαν ἀξίωσιν) in words (τῶν ὀνομάτων)’.Footnote ⁵¹ We can thus translate the stasis-passage as follows (3.82.4):

In the stasis-passage, τῶν ὀνομάτων perhaps refers to valuations and acts of moral assessment that occur in political speeches, political discussion and political undertakings. Such acts of moral assessment are what the examples depict. Although words are not at issue in some examples, Thucydides must assume that the acts of valuations he describes in those examples find linguistic expression. We can read Ex.1 as follows:

The subject matter of the stasis-passage is valuation or moral assessment. The line is committed to valuation finding expression in words, and the term ‘words’ here specifies what sort of valuation is at stake. Similarly, Ex.1 merely assumes the linguistic expression of valuation and its subject matter is an example of valuation. Neither the stasis-passage nor Ex.1 and the following examples are about words themselves and linguistic expressions, let alone about semantics, meaning, or reference. They all are about valuation.

(2.3)

Let us finally flesh out both the stasis-passage and the first example. As we have seen, these passages involve assumptions imported from the preceding context as well as assumptions related to the expression τῶν ὀνομάτων:

According to these lines: (1) Action types remain the same in peace and in stasis. Reckless audacity is reckless audacity in peace as well as in stasis. (2) However, people morally assess or valuate the same action types differently depending upon whether they are in peace or in stasis. In peace, people morally assess reckless audacity normally as reckless audacity. But in stasis, they morally assess reckless audacity abnormally as courage in the protection of partisan interests. (3) Consequently, people describe the same action types differently depending upon whether they are in peace or in stasis. In brief: (1) Reckless audacity is reckless audacity in peace and in stasis. (2) People's moral assessment of reckless audacity in peace is different from their assessment of the very same type of action in stasis. And (3) their description of reckless audacity in peace is different from their description of the very same type of action in stasis.

Two things drift from peace to stasis, moral assessment and the words chosen to describe types of actions. Moral assessment drifts from normal to abnormal; word choice varies from choices that match normal assessments to choices that match abnormal assessments. Two things, however, remain the same in peace and in stasis: action types and words themselves. Indeed, since action types remain the same, we can use them as a criterion to tell that moral assessment and word choice vary between peace and stasis. On the other hand, words themselves, their meaning, do not appear to vary. Indeed, to understand what Thucydides describes, the reader does not need to assume an analysis of words into their phonetic/written marks and their meaning or reference. Moreover, the text is also committed to the possibility that people in stasis are prepared to defend their views, to defend their description of reckless audacity as ‘courage in the protection of partisan interests’, even to argue that it would be wrong to describe reckless audacity as ‘reckless audacity’.Footnote ⁵² Any variation in word meaning is out of the question.

The stasis-passage implies that the way in which people in stasis assess and describe actions is perverse. This way of assessing and describing actions is perverse not merely because, against common sense and the weight of evidence, people in stasis obstinately insist that what is right is wrong and what is wrong is right. This way of assessing and describing is perverse also because people in stasis use the wrong vocabulary to talk about what is right and what is wrong. The question then arises what in particular is perverse and unsettling about assessing reckless audacity as courage in the protection of partisan interests and calling it ‘courage in the protection of partisan interests’?

Reckless audacity is a type of disposition for action which consists in being irrationally overconfident. A consequence of such overconfidence is failing to assess the circumstances and risks in actions. And failing to assess the circumstances and risks eventually prevents people from doing what is morally and practically best. ‘Reckless’ in ‘reckless audacity’ is perhaps an intensifier. The expression ‘reckless audacity’ (τόλμα ἀλόγιστος) is redundant, since audacity is by itself reckless or irrational. Courage, by contrast, is a type of disposition for action which consists in rational confidence, which allows agents to assess the circumstances and risks associated with different courses of action despite their fear; this in turn eventually leads to accomplishing what is morally or practically best. Regarding reckless audacity as courage and calling it ‘courage’ confuses not only right and wrong but also rational and irrational. Thucydides, however, has something more nuanced in mind.

Thucydides is considering not just courage but courage in the protection of partisan interests. While courage is a positive disposition all things considered, courage with the qualification ‘in the protection of partisan interests’ need not always be positive. Courage in the protection of partisan interests is beneficial to a fraction of society, the party and its members; but it may be harmful to the state as a whole. This last point is crucial because democracy demands loyalty to the city-state and its laws over loyalty to a faction, the family, selfish interests and foreign interests.

Thucydides is suggesting that in stasis people place party interests before those of the city-state. This inversion of values and priorities must be part of what he finds perverse. It is also perverse to insist that the right description of ‘failing to assess risks that impend the city-state owing to overconfidence and prioritizing party interests over state interests’ is not ‘reckless audacity’ but ‘courage in the protection of partisan interests’. It is redundant to talk of ‘reckless audacity’, because audacity is always reckless; but it is not redundant to speak of ‘courage in the protection of partisan interests’, because courage is a virtue and courage in the protection of partisan interests is a vice.Footnote ⁵³

Thucydides is diagnosing a pathology which consists in a collective impairment of moral judgement in political matters. We may find it hard to fully appreciate why he finds such impairment scandalous. A close-to-home example would be illuminating. Here is an example involving personal relationships: the loving gaslighting partner. This individual perfectly understands what love and abuse are. Yet they are sincerely convinced that they genuinely love their victim. They are convinced they treat their victim lovingly, and they call their treatment ‘passionate love’, ‘complicated love’, or ‘tough love’. They are even prepared to argue that they are not an abuser but a loving partner and that we should call them a ‘loving partner’. This pathology is perverse by all accounts. Thucydides brings home something similarly unsettling at a collective level.Footnote ⁵⁴ In stasis, people correctly understand what reckless audacity is. Yet they sincerely take cases of reckless audacity to be courage in the protection of partisan interests and describe them as such.

(2.4)

Let us conclude with how we should appreciate the stasis-passage. The passage is not about meaning, semantic drift, or word change but about drift in moral assessment. Word choice, however, does come into consideration. Difference in moral assessment implies difference in word choice in the description of actions. But difference in word choice does not imply any change in the meaning of words. Since the passage is not about any change in meaning and describes a collective spontaneous pathology, it cannot allude to the deliberate creation of new words and meanings characteristic of authoritarian regimes.

Although he denies that the passage is about meaning, Wilson misleadingly uses an example from the USSR to elucidate Thucydides’ point.Footnote ⁵⁵ The USSR had a policy of describing and treating political dissidents as mentally ill,Footnote ⁵⁶ which was a deliberate and systematically planned act of oppression and political propaganda. Its organizers were not sincere when they described and treated political dissidents as mentally ill.Footnote ⁵⁷ Thucydides, however, is talking about an unconscious, unintentional and spontaneous collective condition which leads people to value sincerely what is right as wrong and what is wrong as right.Footnote ⁵⁸ His point is about the obstruction of correct moral judgement in a society which can, in principle, judge correctly. The authoritarian, manipulative and oppressive linguistic practices of the USSR are not a proper cultural and historical context for the stasis-passage. It would be more accurate to compare this passage with antecedent and subsequent Greek views on human flaws, rational impairment and lack of self-control.Footnote ⁵⁹

Homer regarded ἄτη as a certain negative godly or daemonic external force which drives humans out of their senses making them commit terrible irreparable mistakes (Il. 19.85–9, 2.110–15, 8.236–40; Od. 12.371–3, 21.299–302). Aeschylus had a secularized view of ἄτη as mental aberration (Aesch. Sept. 312–16).Footnote ⁶⁰ Early in his career, Plato shared a Socratic view, according to which human flaw results from lack of knowledge (Prt. 345d6–346b8, 352a8–359a1), though he later explains moral failure as resulting from internal conflict among parts of the soul (Resp. 437b7–441c3). For Aristotle moral failure results from internal psychological pressure that prevents humans from firmly grasping that a particular action falls under a general rule (Eth. Nic. 1146b35–1148a11; De motu an. 701a7–29).Footnote ⁶¹

Unlike all the authors quoted above, however, Thucydides specifically and explicitly considers collective impairment of moral judgement, which may explain his introduction of collective phenomena such as words, descriptions and language. But like Aeschylus, Plato and Aristotle—and unlike Homer—Thucydides produces a rational explanation centred in human beings. Like Aristotle, he seems to think that psychological pressure obstructs knowledge and good judgement, though he attributes such psychological pressure to external circumstances such as war, danger and need.

The reading of Thuc. 3.82.4 that I propose preserves consistency between these passages and their preceding and following textual context. It also inserts the passage in a broader intellectual and cultural context of reflection on human flaw. And perhaps more importantly it extracts from Thucydides a valuable observation on human nature: violence, misery, danger, calamity and fear impair our thinking and moral strength both individually and collectively.