Grammatical cases are familiar to us as the forms taken by declinable parts of speech according to their function in a sentence. Historians of grammar and linguistics credit the Stoics with having established the term πτῶσις as technical vocabulary for what we recognize today as the grammatical case, as well as with having advanced a doctrine of five cases: the nominative (ὀρθή) as well as the genitive (γενική), dative (δοτική), accusative (αἰτιατική) and vocative (κλητική) as the oblique (πλάγιαι) cases.Footnote 1 Despite the absence of an established vocabulary as part of an explicit doctrine of cases, the use of πτῶσις to designate a declined word form can be found as early as in the Aristotelian corpus, in which there is evidence that Aristotle was aware of the declension of names and nouns, and perhaps even had a latent theoretical notion of grammatical cases. This much is suggested by his discussion of the distinction between names (ὀνόματα) and cases of names (πτώσεις ὀνομάτων) at On Interpretation (Int.) 16a32–b1, which is taken by later ancient commentators, such as Ammonius, as an indication that Aristotle did not consider the nominative to be a case.
The On Interpretation passage is the Aristotelian locus of a centuries-long debate in the history of grammatical theory that attempts to address a contradiction inherent in the status of the nominative as a case—an oddity which was, according to a scholiast on Dionysius Thrax, first noticed by the second-century grammarian, Apollonius Dyscolus.Footnote 2 As the verbal noun from πίπτειν—‘to fall’—the most literal translation of πτῶσις is a ‘falling’. One view in ancient grammars is that an oblique case—a πλαγία πτῶσις, literally a ‘sideways falling’—is a case because it ‘falls from’ the nominative.Footnote 3 Naturally, then, one might wonder whether the nominative case—the ὀρθή or εὐθεῖα πτῶσις—is also itself a case or a falling. The contradiction, so it goes, arises from the fact that the nominative case, as a πτῶσις, is ‘falling down’, while, as ὀρθή or εὐθεῖα, it is ‘upright’ or ‘standing up’. The nominative case is thus an ‘upright falling’ or a ‘standing-up falling-down’. Ancient grammarians find themselves vexed by how something that is falling down can, at the same time, be standing up. Denying that it is a case is one way to dissolve the contradiction. But if one grants that the nominative is indeed a case, one is then on the hook for explaining what it falls from.
The ancient commentators attribute the former view to Aristotle, despite the absence from his corpus of the established technical vocabulary for the nominative that incites this problem in the first place. Nor is there sufficient reason to believe that Aristotle himself would have been aware of the problem. For at this point in antiquity there is no proper terminology for naming the individual cases—these were coined later. A further wrinkle is that in two other discussions of cases, in the Prior Analytics (An. pr. 48b35–49a5) and the Poetics (Poet. 1457a21), Aristotle lists the nominative form of a word as an example of a case. Against what we find in the On Interpretation, and against the view attributed to him in later antiquity, this strongly suggests that Aristotle indeed considers the nominative to be a case, at least under certain conditions.
This paper offers a reading of the nominative case in Aristotle that renders these passages consistent, by extracting a theory of grammatical case latent in the corpus. It starts with an overview of the problem of the nominative case and examines the view of the nominative case attributed to Aristotle by Ammonius. It then studies passages in the Aristotelian corpus on cases in general and on the nominative case in particular, surveying attempts to reconcile Aristotle’s disparate characterizations of the nominative in relation to casehood. In place of these competing interpretations, I argue that two senses of πτῶσις or case are operative across the corpus—one broad view of cases, on which any morphological change a word can undergo counts as a case, and one narrow view of cases, on which only declined forms of nouns or names count as cases. This helps reconcile the two seemingly opposing characterizations of the nominative in a manner more consistent with what we find in the texts, while also explaining without dismissing the interpretation of Aristotle’s view of the nominative in later ancient commentaries.
1. AMMONIUS ON THE PARADOX OF THE ‘STANDING-UP FALLING-DOWN’ IN ARISTOTLE’S ON INTERPRETATION
Philosophers and grammarians attempt to dissolve the paradox of the nominative case as a ‘standing-up falling-down’ in various ways. An exposition of the problem and two possible solutions to it are presented in Ammonius’ Commentary on Aristotle’s On Interpretation, 42.30–43.20 in the form of a dialogue between representatives defending each of the two solutions. The passage is worth considering in full:
(a) περὶ τῆς κατ’ εὐθεῖαν γινομένης τῶν ὀνομάτων προφορᾶς εἴωθε παρὰ τοῖς παλαιοῖς ζητεῖσθαι πότερον πτῶσιν αὐτὴν προσήκει καλεῖν ἢ οὐδαμῶς, ἀλλὰ ταύτην μὲν ὄνομα ὡς κατ’ αὐτὴν ἑκάστου τῶν πραγμάτων ὀνομαζομένου, τὰς δὲ ἄλλας πτώσεις ὀνόματος ἀπὸ τοῦ μετασχατισμοῦ τῆς εὐθείας γινομένας. τῆς μὲν οὖν δευτέρας προΐσταται δόξης ὁ Ἀριστοτέλης, καὶ ἕπονταί γε αὐτῷ πάντες οἱ ἀπὸ τοῦ Περιπάτου, τῆς δὲ προτέρας οἱ ἀπὸ τῆς Στοᾶς καὶ ὡς τούτοις ἀκολουθοῦντες οἱ τὴν γραμματικὴν μετιόντες τέχνην. λεγόντων δὲ πρὸς αὐτοὺς τῶν Περιπατητικῶν ὡς τὰς μὲν ἄλλας εἰκότως λέγομεν πτώσεις διὰ τὸ πεπτωκέναι ἀπὸ τῆς εὐθείας, τὴν δὲ εὐθεῖαν κατὰ τίνα λόγον πτῶσιν ὀνομάζειν δίκαιον ὡς ἀπὸ τίνος πεσοῦσαν; δῆλον γὰρ ὅτι πᾶσαν πτῶσιν ἀπό τινος ἀνωτέρω τεταγμένου γίνεσθαι προσήκει, ἀποκρίνονται οἱ ἀπὸ τῆς Στοᾶς ὡς ἀπὸ τοῦ νοήματος τοῦ ἐν τῇ ψυχῇ καὶ αὕτη πέπτωκεν· ὃ γὰρ ἐν ἑαυτοῖς ἔχομεν τὸ Σωκράτους νόημα δηλῶσαι βουλόμενοι, τὸ Σωκράτης ὄνομα προφερόμεθα· καθάπερ οὖν τὸ ἄνωθεν ἀφεθὲν γραφεῖον καὶ ὀρθὸν παγὲν πεπτωκέναι τε λέγεται καὶ τὴν πτῶσιν ὀρθὴν ἐσχηκέναι, τὸν αὐτὸν τρόπον καὶ τὴν εὐθεῖαν πεπτωκέναι μὲν ἀξιοῦμεν ἀπὸ τῆς ἐννοίας, ὀρθὴν δὲ εἶναι διὰ τὸ ἀρχέτυπον τῆς κατὰ τὴν ἐκφώνησιν προφορᾶς. ἀλλ’ εἰ διὰ τοῦτο, φασὶν οἱ ἀπὸ τοῦ Περιπάτου, τὴν εὐθεῖαν πτῶσιν ἀξιοῦτε λέγειν, συμβήσεται καὶ τὰ ῥήματα πτώσεις ἔχειν καὶ τὰ ἐπιρρήματα τὰ μηδὲ κλίσεων ἀνέχεσθαι πεφυκότα· ταῦτα δὲ ἐναργῶς ἄτοπα καὶ ταῖς ὑμῶν αὐτῶν παραδόσεσι μαχόμενα. διὰ ταῦτα μὲν οὖν τὴν Περιπατητικὴν περὶ τούτων διάταξιν προτιμητέον.Footnote 4
Concerning the utterance of names in the nominative (eutheian), the ancients used to investigate whether it was proper to call this a case, [or whether] it was itself rather the name of each thing which it names, while the other cases of the name came about by the reshaping of the nominative. Aristotle defends the latter view, and all the Peripatetics follow him, while the Stoics (and since they follow them, those who pursue the art of grammar) defend the former view. But when the Peripatetics say to them, ‘While we are correct in calling the others “cases” because of their having fallen from the nominative, by what reasoning is it justified to call the nominative a case, as though it has fallen from what? (For clearly, it is proper for every case to come from something placed above it)’, the Stoics reply: ‘It too has fallen, namely from the thought (apo tou noêmatos) in the soul; for when we want to make clear the thought (noêma) of Socrates which we have in ourselves, we utter the name (onoma) “Socrates”. So, just as a pen released from a height and lodged upright is said both to have fallen and to have the upright fall, in the same way do we assert that the nominative has fallen from the conception (apo tês ennoias), and that it is upright because it is the archetype of the utterance used in the expression.’ ‘But,’ the Peripatetics say, ‘if you claim that you call the nominative a case because of this, then verbs will also turn out to have cases, along with adverbs, which are not even such as to have inflections (kliseôn)—but this is clearly absurd and at odds with your own teachings. So, for this reason, the Peripatetic arrangements concerning these matters are to be preferred.’Footnote 5
On one side of the debate concerning whether the nominative is a case we find the Stoics and the Hellenistic grammarians, and on the other side are the Peripatetics who follow Aristotle.
One strategy by way of responding to the paradox is to accept that the nominative is a case and to explain why something’s being an ‘upright falling’ does not constitute a vicious contradiction. This is the strategy of the Stoics and of the Hellenistic grammarians who are influenced by them. Because it is not directly relevant for the purposes of this paper, I set aside the details of the Stoic view of cases and flag here only that the version of the Stoic view presented by Ammonius responds to the problem by claiming that the nominative falls from the thought of the object denoted by the word. Hence, the nominative case ‘Socrates’ falls from the thought of Socrates in the speaker’s mind. Because the nominative is the archetypal form of a word, presumably the oblique cases then ‘fall from’ the nominative as modifications of it. As such, the nominative and the oblique cases are indeed both cases, but they are ‘fallings’ in different ways.
Another strategy is to deny that the nominative is a case, so that as the archetypal form of a word it is only upright and not a falling. While the nominative does not fall from anything, the genitive, dative, accusative and vocative are cases because they fall from the nominative. According to Ammonius, this is the preferred strategy of the Peripatetics, who take themselves to be heirs or defenders of the views held by Aristotle.
Text (a) occurs in Ammonius’ discussion of Int. 16a32–16b1. Ammonius is reading the following stretch of Aristotle’s text as an articulation of the view that the nominative is excluded from casehood:
(b) τὸ δὲ Φίλωνος ἢ Φίλωνι καὶ ὅσα τοιαῦτα οὐκ ὀνόματα ἀλλὰ πτώσεις ὀνόματος. λόγος δέ ἐστιν αὐτοῦ τὰ μὲν ἄλλα κατὰ τὰ αὐτά, ὅτι δὲ μετὰ τοῦ ἔστιν ἢ ἦν ἢ ἔσται οὐκ ἀληθεύει ἢ ψεύδεται, τὸ δ’ ὄνομα ἀεί, οἷον Φίλωνός ἐστιν ἢ οὐκ ἔστιν· οὐδὲν γάρ πω οὔτε ἀληθεύει οὔτε ψεύδεται.Footnote 6
‘Philo’s’, ‘to Philo’ and the like are not names (onomata) but cases of the name (ptôseis onomatos). The same account holds for these [as for names], except that [a case], when combined with ‘is’, ‘was’ or ‘will be’ is not true or false, whereas a name always is. Take, for example, ‘Philo’s is’ or ‘Philo’s is not’—so far, there is nothing true or false.
In (b), Aristotle distinguishes between names (onomata) and the cases of a name (ptôseis onomatos), claiming that the genitive and dative forms of ‘Philo’ are not names but cases of a name. To make the distinction apparent, we need an account of what a name is. Earlier in the On Interpretation, Aristotle defines a name as ‘a spoken sound significant by convention, without time, none of whose parts is significant in separation’.Footnote 7 In text (b), this account, he says, holds for cases as well, with the exception that, while names form a truth apt sentence (that is, a sentence that can be true or false) when arranged with the verbs ‘is’, ‘was’ or ‘will be’ (taken existentially), cases of names do not. For example, ‘Philo’s is’ does not form a truth apt sentence, whereas ‘Philo is’ does. This is because ‘Philo’s’ is a case of a name, while ‘Philo’ is a name and not a case. Call this the truth test for names: if a noun forms a truth apt sentence when combined with a form of εἶναι, it counts as a name and not as a case. If a noun fails the truth test (that is, it fails to form a truth apt sentence when combined with a form of εἶναι), it counts as a case and not as a name. Hence the passage establishes that counting as a name and counting as a case are mutually exclusive features of nouns.
One way to read (b) is to understand a ‘name’ or an onoma as the nominative form of a word, since that is the form that helps construct a sentence whose truth value can be evaluated when combined with forms of εἶναι. The nominative passes the truth test, and so cannot be a case. ‘Cases of names’ are then what we recognize as oblique cases, since these fail the truth test in so far as they do not construct a truth apt sentence when combined with a form of εἶναι. This is how the Peripatetics to whom Ammonius refers in (a) interpret Aristotle’s remarks in (b)—as claiming that the nominative, since it is a name, is not a case, and although only the genitive and dative forms appear in (b), the phrase καὶ ὅσα τοιαῦτα in the first sentence of the quoted text is clearly meant to include the accusative and the vocative, since they also fail the truth test and thereby count as cases and not as names.
What we have here, then, are the technical notions of an onoma and of a ptôsis. An onoma is the archetypal form of a noun or the name of the word, which incidentally has the nominative form. To use the terminology of contemporary linguistics, the onoma is the reference or lemma form of the word, which happens to coincide with its nominative form. We also have a technical notion of a case or ptôsis as the form of a name in one of the oblique cases—that is, the genitive, dative, accusative and vocative cases.
2. COMPETING ACCOUNTS OF CASE IN ARISTOTLE
This story of cases in the Aristotelian corpus is complicated by the fact that text (b) is not the only discussion involving ptôsis. At Poet. 1457a19–23, Aristotle offers the following definition of case:
(c) πτῶσις δ᾿ ἐστὶν ὀνόματος ἢ ῥήματος ἡ μὲν τὸ κατὰ τούτου ἢ τούτῳ σημαῖνον καὶ ὅσα τοιαῦτα, ἡ δὲ τὸ κατὰ τὸ ἑνὶ ἢ πολλοῖς, οἷον ἄνθρωποι ἢ ἄνθρωπος, ἡ δὲ κατὰ τὰ ὑποκριτικά, οἷον κατ᾿ ἐρώτησιν ἢ ἐπίταξιν· τὸ γὰρ ἐβάδισεν ἢ βάδιζε πτῶσις ῥήματος κατὰ ταῦτα τὰ εἴδη ἐστίν.Footnote 8
A case of a noun or verb is when the word signifies ‘of’ or ‘to’ a thing, and so on, or for one or many, for example ‘men’ or ‘man’, or it may consist in the mode of utterance, for example in question or command: ‘Did he walk?’ and ‘Walk!’ are cases of the verb ‘to walk’.
In this text, number is also a ptôsis for nouns. Further, we are told, verbs have cases—here, the mood is called a ptôsis. That there are cases not only of nouns but also of verbs aligns with Aristotle’s reference to the ptôsis of a verb at Int. 17a9. At Cat. 1a12–15 and Top. 148a10, etymological derivatives of certain words—for example ‘grammarian’ from ‘grammar’, ‘courageous’ from ‘courage’, ‘beneficially’ from ‘beneficial’—are called ptôseis. In addition to nouns and verbs, at Top. 106b29 and 133b36, the gender and adverbial form of an adjective are ptôseis of the adjective.
If we think of ptôsis as grammatical case, the above texts indeed seem to classify odd categories of words and their features as cases. For example, we would not ordinarily think of a verb as having cases, since verbs are conjugated and not declined. Nor would we think of number or etymological derivatives as cases of nouns. The lack of an established technical vocabulary for grammatical elements contributes to this haziness. Aristotle does not have at his disposal the established language of the conjugation of verbs or the declension of nouns. He does not refer to the nominative case, for instance, as the ὀρθὴ πτῶσις or the εὐθεῖα πτῶσις.Footnote 9 He also does not refer to an ‘oblique’ or πλαγία πτῶσις, nor does he mention any of the grammatical cases using any specific terminology. Rather, as we have seen, he uses an example of a word in the form of a particular case to refer to the case itself.
The absence of an established vocabulary for grammar raises a question about whether the term ptôsis has a consistent denotation across all of its uses by Aristotle, or whether Aristotle uses this term to pinpoint different grammatical phenomena at different points in the texts. This puzzle is a pressing one for our purposes. For while text (b) is indeed the reference point for Aristotle’s view of the nominative case as it is later explicated by Ammonius in (a), the above passage from the Poetics lists the nominative form of ἄνθρωπος as an example of a case—specifically, as an example of number as a case. This suggests that, even if names are not themselves cases, they can have a case, if we construe case as also including features of words such as number. This is compatible with the account at (b). For ἄνθρωπος is a case qua a word in the singular—not qua a word in the nominative. So far, then, we find unusual uses of ptôsis that seem to refer to features of words other than declension, but there is no explicit conflict in the account. Other passages in the corpus, however, such as the following passage from An. pr. 48b35–49a5, do seem to generate exactly this problem, in so far as they suggest that the nominative indeed counts as a case:
(d) ὅρους μὲν γὰρ θετέον καιρὸν καὶ χρόνον δέοντα καὶ θεόν, τὴν δὲ πρότασιν ληπτέον κατὰ τὴν τοῦ ὀνόματος πτῶσιν. ἁπλῶς γὰρ τοῦτο λέγομεν κατὰ πάντων, ὅτι τοὺς μὲν ὅρους ἀεὶ θετέον κατὰ τὰς κλήσεις τῶν ὀνομάτων, οἷον ἄνθρωπος ἢ ἀγαθὸν ἢ ἐναντία, οὐκ ἀνθρώπου ἢ ἀγαθοῦ ἢ ἐναντίων, τὰς δὲ προτάσεις ληπτέον κατὰ τὰς ἑκάστου πτώσεις· ἢ γὰρ ὅτι τούτῳ, οἷον τὸ ἴσον, ἢ ὅτι τούτου, οἷον τὸ διπλάσιον, ἢ ὅτι τοῦτο, οἷον τὸ τύπτον ἢ ὁρῶν, ἢ ὅτι οὗτος, οἷον ὁ ἄνθρωπος ζῷον, ἢ εἴ πως ἄλλως πίπτει τοὔνομα κατὰ τὴν πρότασιν.Footnote 10
As terms, one must posit ‘the right moment’ and ‘requisite time’ and ‘god’, but the premise must be taken in accordance with the case of the name (ptôsin onomatos). For this we say generally about all instances, that the terms must always be set out with the nominative of the names (klêseis tôn onomatôn), for example ‘man’ or ‘good’ or ‘contraries’, not ‘of man’, ‘of the good’ or ‘of contraries’, but the premises must be taken in accordance with the cases (ptôseis) of each. For example, one of two things might be ‘to this’, as equal, or ‘of this’, as the double, or ‘this’, as what is hit or seen, or ‘this’ as in man is an animal, or some other way in which the name (tounoma) falls (piptei) in the premise.
In the final sentence of (d), the nominative form of ἄνθρωπος is listed as an example of a term taken in accordance with its ptôsis as part of a premise. Here, unlike in the Poetics passage in (c), the relevant feature of the word is not its number but its nominative form. This would suggest that counter to his claim in the On Interpretation passage in (b) (and the interpretation of that claim by Ammonius and the Peripatetics to whom he refers) the nominative is indeed a case for Aristotle.
To square the claim in the On Interpretation passage with the suggestion that the nominative is a case in text (d), and thus to vindicate Ammonius’ interpretation of Aristotle’s view of the nominative, throughout the rest of this paper, I will show that the discrepancy comes down to a difference in the relevant sphere of explanation.
3. CASES, NARROW AND BROAD
In my discussion of the On Interpretation text in (b) above, I explained that Aristotle describes a technical notion of both a name and a case, where names are the archetypal forms of nouns, and cases of names are the oblique cases that ‘fall from’ the archetypal form. In texts (c) and (d), Aristotle employs a different notion of ptôsis.
The context of the Prior Analytics passage is Aristotle’s discussion of constructing a syllogism.Footnote 11 The ‘setting out of terms’ in (d) is the process by which each variable in a syllogism is assigned a word or an expression. Aristotle claims that the terms of an argument must be listed according to the nominative form of the name. Presumably, this is because, as Striker explains, ‘the list serves to indicate the things about which an argument is to be constructed’.Footnote 12 Primavesi proposes that part of the utility in setting out the terms in the nominative is that the ordinary form of a sentence does not always correspond to the standard language of a premise within a syllogism.Footnote 13 Sometimes, he claims, it will be more natural for a word in the genitive, for example, to serve as the subject term of a sentence. In these instances, setting out the terms in the nominative will make it easier to convert the sentence into the proper ‘A belongs to B’ formula required for the premise.Footnote 14 The importance of the proper setting out of terms was discussed by Aristotle earlier at An. pr. 48a1–27. In particular, using the wrong expression for a term is one way in which a syllogism can turn out to be fallacious.
Since the archetype of the name happens to have the nominative form, setting out the terms in the nominative amounts to assigning words to the variables of an argument by using their proper linguistic forms. When the terms are then used to construct the premises of the argument, however, they must appear in the appropriate ptôsis—that is, in accordance with how the term functions grammatically in the particular sentence situated within the syllogism. In (d), Aristotle lists examples of relational terms that take a complement in the genitive, accusative and dative cases. The final example is of a non-relational term in the nominative. Clearly, then, when he refers to the ptôsis of the name in (d), he refers to names in the form of the nominative, genitive, accusative and dative cases.Footnote 15 To reiterate the puzzle, that Aristotle lists the nominative as one example of a ptôsis in the Prior Analytics seems to oppose Aristotle’s own comments in (b) as well as the later commentary on (b) in (a).
There have been two attempts to make sense of why Aristotle seems to have conflicting views of the nominative across these texts. Primavesi offers the passing suggestion that in the On Interpretation, the sense of ptôsis is a modification of a basic form of a word, with the basic form of a noun being the nominative singular and the basic form of a verb being the third-person singular active indicative.Footnote 16 He maintains that Aristotle’s account of ptôsis is consistent across the corpus, since modification is appropriate to a word only when it enters a sentence.Footnote 17 Hence Primavesi argues that for Aristotle the nominative is a ptôsis when it is part of a sentence. When a word is not part of a sentence—for example in the process of setting out the terms of a syllogism—it appears as the unmodified basic form, which is not a ptôsis but (at least for nouns) happens to have the nominative form. This reading is based on the final sentence of (d), in which the ptôseis are glossed as ‘ways in which names fall in a premise’. Note, however, that the discussions of names and ptôseis in both (b) and (d) involve sentence formation. In (b), the truth test requires combining a word with forms of εἶναι, thereby constructing sentences out of names that are emphatically not cases. In fact, according to the truth test, it is cases that fail to construct syntactically sound sentences, whereas names do successfully construct such sentences. If Primavesi is right, we should expect names to be ptôseis in virtue of appearing in the sentential context proper as the subject terms of forms of εἶναι. However, counter to Primavesi’s reading, it is the name (and not a ptôsis), which enters the sentence, and the ptôseis, which appear on their own, that are external to the sentence. Further, the different forms of ἄνθρωπος that are listed as cases in (c) occur outside of any sentential context. Hence, the presence or absence of the sentential context cannot solve this puzzle, and the reading makes too much of the comment at the end of (d).
Thorp similarly claims that, although names and the nominative case are identical in form, they are different in essence owing to the fact that cases are the forms in which names turn up when they appear in a sentence.Footnote 18 He points out that in the On Interpretation passage, Aristotle only provides the genitive and the dative as examples of ptôseis. Thorp claims that, because the nominative is not the only case missing from (b), Aristotle’s expression of his own view is misleading, and not in fact inconsistent with (c) and (d), for he must have also intended to include the nominative as part of the elliptical expression καὶ ὅσα τοιαῦτα at the beginning of (b). This interpretation suffers from a similar problem described above—namely, that the source of the discrepancy cannot be the presence or absence of the sentential context. But, in addition, Thorp asserts that his interpretation of these texts implies that subsequent centuries worth of readings, including those of Ammonius and the Peripatetics in (a), are simply mistaken about Aristotle’s view of the nominative. There is certainly a broader question about how we should understand the contributions of the ancient commentators as they pertain to interpreting the original texts, but it is not sufficient to dismiss an entire tradition without addressing why they might have read Aristotle this way, apart from the possibility that they might have anachronistically imposed on the corpus a response to a problem which only rears its head decades after Aristotle.
Rather, the discrepancy is a matter of there being two readings of ptôsis at work across these texts. Consider the notion of a ptôsis we encountered in the Poetics, the Categories and the Topics as the number of a noun, as the mood of a verb, as the etymological derivative of a certain word, and as the gender and adverbial form of an adjective. This seems to be a broader notion of ptôsis as the more general grammatical phenomenon of inflection. ‘Inflection’ means any morphological change that any type of word can undergo. For example, the tense, mood and person of a verb are ptôseis in the sense that they are all morphological changes that the basic form of a verb can undergo, with or without occurring in a sentence. The notion of ptôsis as inflection in general, however, does not seem to be what Aristotle has in mind at the On Interpretation passage in (b). There, ptôsis refers only to what we call the oblique cases; notably, while the nominative is a ptôsis in (d), it does not count as a ptôsis in (b)—rather, it is the onoma, as the archetype of the noun from which the (oblique) cases fall. This suggests a second, more restricted understanding of ptôsis, as a change an onoma (understood not merely as a ‘word’ or ‘noun’ but as a name—that is, the archetype of a noun) can undergo.
How does this reconcile the seemingly disparate views of the nominative across these texts? Let us consider the implications of construing ptôsis as referring both to the general phenomenon of inflection and to what happens to an onoma when it ‘falls’. First, the oblique cases will be ptôseis in both the broad sense and the more restricted sense. They are ptôseis in the broad sense because they are inflections. They are ptôseis in the restricted sense because they are, more specifically, the inflected forms that an onoma, as their archetype, can take. The nominative, on the other hand, is a ptôsis only in the broad sense, because it happens to be the grammatical form of an onoma. But it is not a ptôsis in the more restricted sense because, strictly speaking, it is the archetype from which the (oblique) cases fall—it is not itself a change from anything.
The narrow and broad readings of ptôsis do not therefore track whether the word occurs in the sentential context. The difference, rather, is that each notion of ptôsis belongs to a different sphere of explanation. The context of the claim about ptôseis at (d) is one in which the morphological features of the words in a premise reveal the relations between the terms in the argument. Consider the example of an argument Aristotle provides directly before (d): ὁ καιρὸς οὐκ ἔστι χρόνος δέων· θεῷ γὰρ καιρὸς μὲν ἔστι, χρόνος δ’ οὐκ ἔστι δέων διὰ τὸ μηδὲν εἶναι θεῷ ὠφέλιμον. We list the terms of the argument—καιρόν, χρόνον δέοντα and θεόν—in their nominative forms, but we understand the premises by reference to the ptôsis of each term as it appears in the sentence. When constructing the argument itself, we therefore take θεόν, for instance, in the dative case, because it reveals what this term is doing in the argument and how it logically relates to the other terms. In this example, it is the thing to which καιρός (which, as the subject, appears in the premises in the nominative case) belongs. It is therefore the more general phenomenon of inflection that is relevant to this discussion. For it is inflection, including the inflection of a word in the nominative form, which helps reveal the logical relations between the terms of an argument. Contrast this with the context of (b), in which Aristotle is concerned not with the logical relations between the different terms in an argument, but with offering an analysis of simple statements. Such statements consist of a subject expression, which he says is an onoma (the archetype of a name), and of a predicate expression, which says something about the thing named.Footnote 19 ptôseis, understood here in the narrow sense of the ptôseis of names, cannot have this function because the oblique cases cannot play the role of naming in a simple sentence—only the name, the onoma, can do this.Footnote 20 The contrast between the name and the forms that fall from the name is crucial for this difference in function. It is precisely because the ptôseis are ptôseis of onomata that they cannot play this role, and it is because onomata are names and archetypes that they are uniquely suited to this task. In turn, the nominative term both within a premise and when it is listed in the setting out of terms does not, strictly speaking, play the naming role. Rather, in a premise, it designates what it is that belongs to something else, while, in the setting out of terms, it functions as the standard form of a word that will appear in the syllogism in question.
Whether the nominative is a case, then, does not depend on whether the word occurs in a sentence, for the occurrence of the word in a sentence is common to both names and cases. Rather, the casehood of the nominative comes down to what is being explained. If·the explanandum has to do with the type of word that can name things in the subject position of a simple sentence, then, strictly speaking, only a name will do. If, on the other hand, the explanandum has to do with constructing an argument so as to reveal the logical relations between the terms of an argument, then the term that appears in the nominative case (the grammatical form a name happens to have) will reveal the subject of the argument—the thing to which the other terms somehow relate. So, the nominative is, in general terms, a ptôsis in virtue of the fact that it happens to be the grammatical form of a name. However, strictly speaking, the nominative is not a ptôsis of a name in the same way in which the oblique cases are ptôseis of a name; in the narrow sense of ptôsis, there is only the name and no nominative.
This means that Ammonius, in his commentary in (a), is not in fact misinterpreting Aristotle’s view of the nominative case. It does, after all, appear in Ammonius’ commentary on the section of text in which the narrow sense of ptôsis is operative. But why does he take (b) as Aristotle’s final word on the matter? Two factors are at play.
First, unlike Aristotle, the later commentators have a conventional vocabulary for Greek grammar at their disposal—a vocabulary that is finally able to distinguish between the various types of inflections which for Aristotle are subsumed under the general umbrella of ptôsis, taken broadly. In other words, by the time the Aristotelian corpus reaches Ammonius, there is technical terminology which, for example, differentiates the cases of nouns from the conjugations of verbs, and both of them from how adverbs are related to adjectives. The later commentators must have been reading Aristotle’s view of ptôsis from this intellectual vantage point, filling in more precise grammatical language where it is lacking in Aristotle. This explains why, despite Aristotle’s claims in the Poetics, the Categories and the Topics, the end of (a) depicts the Peripatetics responding to the Stoics by claiming that it is an absurd consequence of the Stoic view that verbs and adverbs have ptôseis. If we deny that there is a broader sense of ptôsis at play outside of the On Interpretation, as the notion of morphological change in general, the Peripatetics would not find themselves in the position to level this charge against their opponents. Otherwise, we would have to commit to the unlikely assumption that they did not have access to these other texts in Aristotle’s corpus. In (a), the relevant type of ptôsis is therefore not the broader phenomenon of words having certain forms (on which the nominative, because it is a word form, is a ptôsis) but the more restricted notion of the ptôsis of a name (on which the nominative is simply the name, and not itself a ptôsis).
Second, the questions about whether the nominative is a case and whether something’s being an upright falling constitutes a vicious contradiction are not problems that Aristotle raises or addresses. Note, for example, that the reason for excluding the nominative from casehood in (b) is not that it is a contradiction to say that something is an upright falling. Rather, it is because the question being addressed has to do with naming and the formation of simple sentences using einai as the verb. The fact that Aristotle does not address whether there is a contradiction inherent in the very notion of the nominative case is due to the fact that he does not have the language that causes this problem—again, the nominative case as an orthê or eutheia ptôsis is not established until later. As such, the contradiction never arises as something to be noticed or addressed—it is a later intervention of Chrysippus and the Hellenistic grammarians. The post-Aristotelian philosophers become interested in this paradox, which is then taken up by the later commentators as a lens for interpreting Aristotle’s claims in the On Interpretation. Aristotle himself is therefore not at any point concerned with the problem of the nominative case, and his reasons for differentiating names from cases are different from the reasons for excluding the nominative from casehood that are provided by Ammonius and the Peripatetics in (a). The narrow notion of a ptôsis is, then, the one that is operative in (a).
4. FINAL REMARKS
Despite Stoic origins of an explicit doctrine of cases, this article has shown that already in Aristotle’s writings there is discussion of an implicit theoretical notion of ptôsis as a grammatical case. Taken broadly, ptôsis refers to the morphological changes and features of words in general, regardless of the part of speech and regardless of whether the word appears in a sentence. The nominative is a ptôsis in this sense, in so far as it is the grammatical form which the archetype of a word happens to have. On a narrower understanding, ptôsis is a form or a modification of an onoma—a modification of a name that can take the form of the genitive, accusative, dative and vocative. The nominative is not a ptôsis in this narrow sense, since in this context the nominative is simply the name itself, and not a modification of anything. Later commentators, who had at their disposal a more comprehensive technical language for grammar, take this to imply that for Aristotle the nominative is not a case. This is because they take Aristotle to be discussing ptôsis in the narrow sense. This is not to say that Aristotle is the first to notice the phenomenon of declension. As Frede notes, ‘to be aware of … declension … is not yet to have a theoretical notion of the grammatical cases’.Footnote 21
One loose thread is why the word ptôsis, which ordinarily denotes a kind of descent, also comes to denote a grammatical phenomenon. The final sentence of (d) is an important clue. To someone observing language in a world before grammatical rules have been articulated, language must seem obviously rule-governed, but by rules not yet articulated in a science of grammar. As such, Aristotle’s remark in (d) that taking a word in accordance with its case is to take it the ‘way in which the name falls in the premise’ suggests that the grammatical sense of ptôsis is something like the falling or ptôsis of dice—that is, governed by an internal rule, our ignorance of which leads us to the language of luck and chance.Footnote 22 Indeed, piptei already has this figurative sense in Aristotle’s time, not merely as spatial descent but as something ‘turning out’. The accusative case is a ptôsis in the sense that it is the form a certain word ‘turns out to have to take on’ in order to appear in a certain sentence as, say, the object of a verb. Or ‘courageous’ is a ptôsis in the sense that it is the form the word ‘courage’ happens to take on in order to function as an adjective.Footnote 23 Hence etymologically ptôsis as a grammatical expression, in so far as it is a figurative extension of the verb piptei, indirectly derives from the image of spatial descent described by Ammonius.
Open access
