2 May 2017

Luhtala ( On the Origin of Syntactical Description in Stoic Logic, “Compound Predicates and the Notion of Action”, summary


by Corry Shores


[Search Blog Here. Index tabs are found at the bottom of the left column.]


[Central Entry Directory]

[Stoicism, entry directory]

[Anneli Luhtala, entry directory]

[Luhtala, Syntactical Description in Stoic Logic, entry directory]


[The following is summary. It redoes this entry. All boldface and bracketed commentary are my own. Paragraph enumerations are also my own, but they follow the paragraph breaks in the text. Please forgive my distracting typos, as proofreading is incomplete.]




Summary of


Anneli Luhtala


On the Origin of Syntactical Description in Stoic Logic


Ch.5 The Stoics


5.5 Stoic Logic



The Component of Meaning (Σημαινόμενα)

Compound Predicates and the Notion of Action





Brief summary:

The Stoics noted compound predicates, which are ones whose syntactical structures (namely, oblique cases) allow for the designation of a patient to the action. It is thus something like a transitive formation. But verbs by themselves are not thought of as signifying actions. Rather, actions are corporeal and pertain just to bodies and thus to nouns. Hence the actions are somehow a part of the agents and patients and are thus built into the nouns signifying these bodies (on the basis of the oblique cases of the nouns, which indicate the agent/patient relations). Verbs (and their predicates) designate the relations holding between these bodies and not the actions involving them.  Many nouns signify actions, as for example the noun “action” itself. So while normally we might find it odd that for the Stoics verbs do not represent actions, in fact this distinction allows nouns and verbs to maintain distinct sorts of senses.









[The domain of sentences of Stoic syntactical studies are ones yielding a perfect sense. Such sentences have either a simple or compound predicate (that is, one that requires the addition of an oblique case to express a perfect sense).]


[Previously we examined the Stoic notion of self-sufficiency (αὐτοτέλεια). It had something to do with whether or not a sayable’s predicate was completed by a subject. But the terminology in this prior section was a bit vague and ambiguous.] Although the Stoic notion of self-sufficiency was defined somewhat vaguely, “it established itself as the standard way of defining the sentence in later grammatical tradition” (88). Syntactical studies take as their domain sentences that yield “a perfect sense” (88). Sentences that can yield a perfect sense may have either a simple predicate, like “Socrates writes”, or a compound predicate, namely, “one which requires the joining of an oblique case to express a perfect sense” (88). [I do not yet know what an “oblique case” is. Here is one definition: “2. In the traditional grammatical descriptions of Latin and Greek, denoting any case form of a noun other than the nominative or vocative” (Trask 194). Originally by “compound predicate”, I thought that it meant two predicates joined by “and” like “snow is white and is cold”. But I wonder if it means something more like adding phrases of other sorts of cases to the predicate, as for example, “he hit me” becoming “he hit me with a stone” or something like that. (From what is said later, it seems the oblique case involves designating a patient to an action, like “he cooks” becoming “he cooks the fish”).]

[There is some textual evidence to support the simple/compound predicate distinction.]


Luhtala then discusses the textual evidence for the simple/compound predicate distinction. We do not see compound predicate in Diogenes Laertius. However, he does define ῥῆμα (verb) as a simple predicate (ἀσύνον κατηγόρημα), which implies there must also be a compound predicate. Luhtala also notes that the Stoics referred to compound predicates as “less than a predicate” (88). We will examine this below.

[The simple/compound predicate distinction can be understood in terms of transitivity, which involves agent and patient and thus an active-passive opposition (as well as the Stoic notion of causation).]


We can make the simple/compound predicate distinction also in terms of transitivity, which “has to do with the expression of action which involves two persons, the agent and the patient, and it is transitive verbs that partake of the active-passive opposition” (89). Luhtala will explain how transitivity is related to Stoic syables and also how it is related to the Stoic theory of causation. Luhtalla is the first to give detailed analysis on these topics, although others have realized “the importance of the notion of action to the Stoic proposition” (89).

[The contents of verbs and predicates were not understood in terms of action, because many activities were thought to be corporeals rather than the actions of bodies.]


Action is clearly an important notion for Stoic conception of a proposition. However, determining this importance on the basis of textual evidence is not entirely straightforward, Luhtala explains (89). She continues, “Given that the Stoics were concerned with action as the contents of the verb/predicate, one might expect them to have defined the verb as signifying action and undergoing of action. This is how the grammarians defined the verb later on”  (89). However, “neither the verb nor the predicate was defined in terms of action by the Stoics” (89). Later Luhtala will argue that Stoics intentionally did not define the verb and predicate in terms of action, because many activities were understood as bodies themselves rather than as the activities of bodies. The verb, however, must signify something incorporeal, for the Stoics.

According to my view, the Stoics did not describe the semantic content of the verb by means of such nouns as action (ἐνέργεια) because various activities (such as walking and dancing) were regarded as corporeal things in the Stoic theory. It is crucial for the Stoic theory that the verb signifies something incorporeal.

(89-90, boldface and underlining mine)

[Even though for the Stoics predicates and verbs express incorporeals and thus not actions, which are corporeals, the Stoics still saw there being transitive verbs and thus predicates that are active or passive (thus seemingly involving corporeal activity).]


[Predicates should express something incorporeal and thus not an action (which is corporeal), yet] nonetheless, the Stoics did deal with active and passive predicates [which suggests they saw the predicates as dealing with agents acting on patients and thus as dealing with actions.] Diogenes Laertius gave examples of active and passive predicates. One sort are verbs of perception, like, ἀκούει (hears) and ἀκούομαι (I am heard). These are transitive verbs involving two people or things and allowing “for active/passive opposition” (90). There are also neuter predicates that are intransitive verbs, like περιπατεῖν (to walk) (90).

[Many scholars think that the neuter, intransitive predicates are the standard sort in Stoic logic, but Luhtala argues that active predicates were important too.]


Luhtala writes that “it is representatives of neuter predicates that predominate in the examples of the Stoic proposition, and we are told that the ‘intransitive’ proposition is the Normfall in Stoic logic (Prantl [1855] 192: 443; Pohlenz 1939: 170; Müller 1943: 50)” (90). Luhtala disagrees and instead argues that “the ‘transitive’ state of affairs probably held equal prominence in the Stoic analysis” (90). The Stoics had a notion of the active predicate as a ‘direct predicate’ (ὀρθὸν κατηγόρημα) which “can easily be understood as the most unmarked predicate” (90). Because active predicates are called “direct”, that means they must have a central importance for the Stoics (90).

[There are many cases of Stoic transitive predicates.]


Luhtala then notes that Diogenes Laertius’ definitions can be “applied to ‘transitive’ constructions” as for example ones “designed to describe states of affairs involving more than one participant ... (πρᾶγμα συντακτὸν περί τινος ἤ τινῶν)” (91). There are other cases as well where Stoic predicates were understood as active or passive [see p.91]; for example: “Α scholiast on Dionysius Thrax points out that it is typical of a verb to indicate a state of affairs (πρᾶγμα); and these states of affairs are accomplished by persons in so far as they act or undergo actions.153 [footnote 153:] Toὐ δὲ ῥήματος ἴδιον τὸ σημαίνειν πρᾶγμα. τὰ δὲ: πράγματα διὰ τῶν ἀνθρώπων κατορθοῦται ἢ ὡς ἐνεργούντων ἢ ὡς πασχόντων (GG Ι.3: 215, 28-29)” (91).

[The verb in a sentence indicates a state of affairs (πρᾶγμα), but the state of affairs is not something entirely exterior to the proposition. It is somehow a part of the proposition.]


[Above we noted the idea that a verb indicates a state of affairs, which is accomplished by persons undergoing actions.] Luhtala accepts this interpretation, but she specifies “however that the state of affairs (πρᾶγμα) is also a feature of the proposition”. [I am not sure I grasp the next points, so I will guess them to  be the following. We might normally think of the state of affairs as being completely distinct from the uttered sentences expressing them. But somehow we might not see this metaphysical distinction. I am not sure how yet, but we should try to build a conception. Let us look first at what Luhtala writes.]

Ι accept this interpretation specifying however that the state of affairs (πρᾶγμα) is also a feature of the proposition.154 The proposition was defined as a complete state of affairs (πρᾶγμα αὐτοτελές) by the Stoics (Diog. Laert. VII,65); it was defined as exhibiting a complete sayable (λεκτόν αὐτοτελές) by Aulus Gellius.155 The component of meaning (σημαινόμενα) was known even as the component of states of affairs (πράγματα, Diog. Laert. VII, 63). States of affairs (πράγματα) and sayables (λεκτά) would actually appear to be parallel notions within this component so that they reflect two different aspects of the same units of sayables, namely of the predicate and proposition.156 The notion of sayable (λεκτόν) points to the kind of existence or rather non-existence associated with our thought | and speech, while the state of affairs (πρᾶγμα) stands for the contents of these units of thought and speech in terms of their signifying states of affairs or action.


154 As far as Ι can see, the Stoic philosophers did not say that the verb signifies action and undergoing action (see p. 142).

155 λεκτόν αὐτοτελὲς ἀποφαντὸν ὅσον ἐ φ’ ἑαυτῷ) (Noct. Att. XVI, 8,1.

156 All these works belong to the department of πράγματα(‘states of affairs’), which | deals with propositions and other kinds of sentences as well as their components, predicates, cases etc. (Diog. Laert. VII, 190).


[We will go step by step now, but I am a little confused by all the terminology. “The proposition was defined as a complete state of affairs (πρᾶγμα αὐτοτελές) by the Stoics (Diog. Laert. VII,65)”.  Let us look at the Diogenes text:

A judgement is that which is either true or false, or a thing complete in itself, capable of being denied in and by itself, as Chrysippus says in his Dialectical Definitions: “A judgement is that which in and by itself can be denied or affirmed, e.g. ‘It is day,’ ‘Dion is walking.’” The Greek word for judgement (ἀξίωμα) is derived from the verb ἀξιοῦν, as signifying acceptance or rejection; for when you say “It is day,” you seem to accept the fact that it is day. Now, if it really is day, the judgement before us is true, but if not, it is false.


A proposition is that which is true or false, or a complete state of affairs which, so far as itself is concerned, can be asserted, as Chrysippus says in his Dialectical definitions.

(Long & Sedley 202d)

Ἀξίωμα δέ ἐστιν ὅ ἐστιν ἀληθὲς ἢ ψεῦδος: ἢ πρᾶγμα αὐτοτελὲς ἀποφαντὸν ὅσον ἐφ᾽ ἑαυτῷ, ὡς ὁ Χρύσιππός φησιν ἐν τοῖς Διαλεκτικοῖς ὅροις "ἀξίωμά ἐστι τὸ ἀποφαντὸν ἢ καταφαντὸν ὅσον ἐφ᾽ ἑαυτῷ, οἷον Ἡμέρα ἐστί, Δίων περιπατεῖ." ὠνόμασται δὲ τὸ ἀξίωμα ἀπὸ τοῦ ἀξιοῦσθαι ἢ ἀθετεῖσθαι: ὁ γὰρ λέγων Ἡμέρα ἐστίν, ἀξιοῦν δοκεῖ τὸ ἡμέραν εἶναι. οὔσης μὲν οὖν ἡμέρας, ἀληθὲς γίνεται τὸ προκείμενον ἀξίωμα: μὴ οὔσης δέ, ψεῦδος.


So somehow a judgment or proposition (written here as ἀξίωμα) can be a complete state of affairs. It is not yet entirely clear to me how they can be equivalent. The best way I have so far to understand this is to think of situations in the world as having a “sense” to them, which is their λεκτόν or sayable. So perhaps a complete state of affairs is one with a λεκτόν and thus with a propositional sense built into it somehow. (See section 5.5.1 and section 5.5.2) The next sentence was: a proposition “was defined as exhibiting a complete sayable (λεκτόν αὐτοτελές) by Aulus Gellius”. So perhaps we can say that both the proposition and the state of affairs corresponding to it share the same sayable / λεκτόν.  Next: “The component of meaning (σημαινόμενα) was known even as the component of states of affairs (πράγματα, Diog. Laert. VII, 63)”. Before we look at the text, recall from section 5.5.2 that σημαινόμενα is more or less the same as the sayable / λεκτόν, only under this terminology (σημαινόμενα) it is understood as  a component of meaning, corresponding to its linguistic expression (σημαίνοντα). Let us again look at the text (I will insert Greek terms based on prior cases, but perhaps inaccurately):

To the department dealing with things as such (πραγμάτων) and things signified (σημαινομένω) is assigned the doctrine of expressions, including those which are complete in themselves, as well as judgements and syllogisms and that of defective expressions comprising predicates (κατηγορημάτων) both direct and reversed.

By verbal expression they mean that of which the content corresponds to some rational presentation.


The topic which deals with states of affairs and significations includes that of sayables, both those that are complete and propositions and syllogisms, and those which are incomplete, and active and passive predicates. (2) They say that a sayable is what subsists in accordance with a rational impression.

(Long & Sedley 296d)

Ἐν δὲ τῷ περὶ τῶν πραγμάτων καὶ τῶν σημαινομένων τόπῳ τέτακται ὁ περὶ λεκτῶν καὶ αὐτοτελῶν καὶ ἀξιωμάτων καὶ συλλογισμῶν λόγος καὶ ὁ περὶ ἐλλιπῶν τε καὶ κατηγορημάτων καὶ ὀρθῶν καὶ ὑπτίων.

Φασὶ δὲ [τὸ] λεκτὸν εἶναι τὸ κατὰ φαντασίαν λογικὴν ὑφιστάμενον.


The idea seems to be that the state of affairs (πρᾶγμα) in the world has a rational sense to it, the λεκτόν, which by means of our senses makes in our mind a rational impression (λογικὴν). When we form a sentence expressing that rational impression, its meaning or sense (σημαινόμενα) is equivalent to the λεκτόν, which composes the rational sense of the state of affairs (πρᾶγμα). Perhaps on this basis we may say that the state of affairs and the sense of the sentence expressing its λεκτόν are equivalent or identical, but I am not sure I have uncovered the exact line of reasoning to do so yet. Next: “States of affairs (πράγματα) and sayables (λεκτά) would actually appear to be parallel notions within this component so that they reflect two different aspects of the same units of sayables, namely of the predicate and proposition.” I cannot decipher this well, but maybe it is the following. The sayable (λεκτόν) is the sense of a state of affairs (πρᾶγμα). But the state of affairs, maybe Luhtala is saying, is just what corresponds to the predicate. So “Dion is walking” would express the λεκτόν and “is walking” expresses the πρᾶγμα. But I am not sure at all yet. Next: “The notion of sayable (λεκτόν) points to the kind of existence or rather non-existence associated with our thought | and speech, while the state of affairs (πρᾶγμα) stands for the contents of these units of thought and speech in terms of their signifying states of affairs or action.” I am not sure how to understand this. Maybe the first idea here is that the sayable (λεκτόν) is incorporeal and thus it is a matter of the incorporeal non-existence of the thoughts and speech expressing that λεκτόν. And maybe the second point is that the contents of thought and speech are the states of affairs expressed by them, and thus “contents” means something like referents. Let me consider another way to think about this before moving on. Let us think of a state of affairs as a factuality somehow. There is a fact in the world, Dion is walking, or maybe just Dion’s walking, if we must stick only with actions or predicates. The statement “Dion is walking” also has a certain factuality about it, namely, its truth. The factuality of Dion’s walking is the same in both the states of affairs in the world and the state of affairs in the sentence. Let me offer yet another interpretation. Let us think of the λεκτόν as  primary or fundamental. It on the one hand is expressed in the physical world as the state of affairs there and on the other hand it is expressed in the sentence corresponding to those states of affairs.]

[The meaning of a sentence (σημαινόμενον), the state of affairs (πρᾶγμα), and the sayable (λεκτόν) are all either one and the same or somehow very near the same.]


Luhtala now examines textual evidence to show how states of affairs (πράγματα) are to be understood as bound up with saybles (λεκτά). Sextus Empiricus equates the meaning of a sentence (σημαινόμενον) with the state of affairs (πρᾶγμα) (92).

The signification (σημαινόμενον) is the actual state of affairs (πρᾶγμα) revealed by an utterance, and | which we apprehend as it subsists in accordance with  our  thought (διανοίᾳ).

(Long & Sedley 196-197, insertions mine) [Note, I did not follow Luhtala’s insertions exactly, because I wanted both σημαινόμενον and πρᾶγμα, where she just has πρᾶγμα of the two, and also she adds ὑπάρχειν for “subsists”, but I was unable to find it in the Greeg text. See below.]

σημαινόμενον δὲ αὐτὸ τὸ πρᾶγμα τὸ ὑπ' αὐτῆς δηλούμενον καὶ οὗ ἡμεῖς μὲν ἀντιλαμβανόμεθα τῇ ἡμετέρᾳ παρυφισταμένου διανοίᾳ

(Sextus Empiricus VIII,11-12).


Diogenes Laertius equates sayables (λεκτά) with states of affairs (πράγματα):

Saying (λέγειν) is different from voicing (προφέρεσθαι). For utterances (φωναι) are voiced but it is states of affairs (πράγματα) which are said – they, after all, are actually sayables. (Diog. Lart. VII,57)


[It is still not entirely clear to me if the states of affairs are somehow to be thought as inherent to the proposition, or simply corresponding to it. From the following wording, it would seem to correspond.] She writes:

According to both authors ‘states of affairs’ subsist in accordance with our thought and are sayable (by means of λεκτά). The parallelism of these notions is reflected in the Stoic terminology, according to which incompleteness and completeness pertain to both sayables (λεκτά) and states of affairs (πράγματα). A proposition can be a complete sayable (λεκτόν αὐτοτελές) and a complete state of affairs (πρᾶγμα αὐτοτελές). A predicate is an incomplete sayable (λεκτόν ἐλλιπές) and, obviously, an incomplete state of affairs, that is, something that can be associated with one or more subjects to form a complete state of affairs.157


157 The intimate relationship betweeη λεκτά (‘sayables’) and πράγματα (‘states of affairs’) in Stoic thought was recognized early οn but this aspect of the proposition is rarely discussed in any detail (see e.g. Mates 1962: 11, 12, n.8 citing Steinthal 1890, 1: 288; Hadot 1978: 314; Rist 1971: 56, n. 11; Long 1971: 77).


[For the Stoics, action has a transitive nature in the sense that the action is directed toward a patient.]


[We noted above in section that for the Stoics, predicates and verbs express incorporeals and thus not actions, which are corporeals. Nonetheless, we also noted, the Stoics still thought that there were transitive verbs and thus predicates that are active or passive (thus seemingly involving corporeal activity). Here Luhtala will explain that for the Stoics, the idea of action also has a transitive nature, and she gives evidence for it on the basis of the notion of actions being “direct” (ὀρθόν).] “Simplicius provides evidence of the ‘transitive’ nature of the Stoic idea of action” (92). Stoic action involves an action directed toward a patient, while Aristotlian action is more general, exemplified by such [intransitive] verbs as ‘to walk’ and ‘to sit’ (92-93). [So “direct” seems like a metaphysical or physical determination meaning that there is an action directed toward the patient. But also, since predicates can also be called direct, that means this conception has a linguistic parallel.] “Furthermore, the fact that such action was labeled direct  (ὀρθόν) suggests that action in a physical context was handled in terms parallel to linguistic terms; the active predicate was known as direct predicate (κατηγόρημα ὀρθόν).” The scholiasts of Dionysius Thrax, referring in all likelihood to the Stoics, also noted the association between action and directness (93).

[For the Stoics, action is a matter of active and patient bodies, with the predicate designating the relation between them. The verb does not however signify action or the undergoing of action.]


[I am not sure I follow the next ideas. They might be the following.] In Stoic physics, “action and undergoing of action were regarded as crucial features of bodies” (93). [This means perhaps that acting and undergoing action were less important for actions themselves. Thus] “the fact that the verb was not defined in the same semantic terms begins to become more understandable” (93). “the Stoics reserved the notion of action to pertain solely to bodies” (93). Now, “Bodies are represented by the nominal parts of speech in the linguistic theory” (93). [Actions are bodily, but verbs do not represent bodily and thus perhaps not some activity but rather somehow an incorporeal sayable.] “and in a striking contrast, the verb is said to signify something incorporeal, a predicate, a sayable which exists only in our thought and speech” (93). [Luthala’s next point seems to be that the predicate, which unites the active and patient bodies, is what gives articulation to the relation between these two bodies.] “The Stoic predicate is a unity of our thought, a relational notion which unites the two bodies involved in action and undergoing of action. As far as I can see, it was impossible for the Stoics, from the ontological point of view, to say that the verb signifies action and undergoing of action” (93).

[Later grammarians rejected the Stoic ontology which called for nouns to signify actions (which pertain to the bodies signified by the nouns) and verbs (which signify the relations holding between such bodies). Grammarians coming much later criticize this move to define verbs as actions, because many  nouns signify actions.]


Later on, grammarians, against Stoic ontology, defined the verb in terms of action and undergoing action (93). But this proved problematic, and Medieval grammarians criticized this move. For, many nouns also signify actions, so it became hard to make the distinction between nouns and verbs.

To say that the verb signifies action and undergoing of action means that it is impossible to differentiate sufficiently between the signification of the noun and the verb , which was always at issue in these definitions. This is because even nouns signify action (e.g. actus), which cannot be a peculiar characteristic of the verb; it thus became impossible to describe the difference between the verb agit (‘acts’) and the noun | actus (‘action’) in terms of signification. The Stoics posited, as far as as Ι can see, a distinction between the different ways in which nouns and verbs signify; the noun signifies a body, while the verb signifies something pertaining to that body.









Luhtala, Anneli. 2000. On the Origin of Syntactical Description in Stoic Logic. Münster: Nodus.



Other texts, cited by Luhtala:


Hadot, Pierre. 1980. “Sur divers sens du mot pragma dans la tradition philosophique grecque”. Aubenc (1980: 309-319).


Diogenes Laertius: Lives of eminent philosophers. Translated by Robert D. Hicks. 2 vols. London: William Heinemann / Cambridge, Mass.: Harvard University Press 1950.

The Perseus Greek page for the Diogenes’ passages:


The Perseus English page for the Diogenes’ passages:



Long, Anthony A. 1971. “Language and Thought in Stoicism”. Long (1971a: 75-113).


Long, Anthony A. 1971a. (ed.) Problems in Stoicism. London: The Athlone Press.


Rist, John M. 1971. “Categories and their Uses”. Long (1971a: 38-57).


Mates, Benson. 1961. Stoic Logic. Berkeley, Los Angeles: University of California Press. [Second edition. Originally published in 1953 as Volume 26 of the University of California Publications in Philosophy.].


Müller, H.-E. 1943. Prinzipien der stoischen Grammatik. Diss. Rostock.


Aulus Gellius: Noctes Atticae. Ed. by Peter K . Marshall. 2 vols. Oxford: Clarendon Press 1990.


Pohlenz, Max. 1939. Die Begründung der abendländischen Sprachlehre durch die Stoa. Göttingen: Vandenhoeck & Ruprecht. (Nachrichten von der Gesellschaft der Wissenschaften zu Göttingen. Philologisch-historische Klasse. Neue Folge. 1.3, Heft 6.).


Prantl, Carl. 1855-70. Geschichte der Logik im Abendlande. Vol 1.  Leipzig: Hirzel. [Repr. Graz: Akademische Druck- und Verlagsanstalt, 1955.].


[GG] Scholia in Dionysii Thracis Artem Grammaticam. GG I.3. Ed. by Alfred Hilgard. Leipzig: Teubner 1878-1910.



Texts I cite:

Trask, R.L. 1993. A Dictionary of Grammatical Terms in Linguistics. London and New York: Routledge.


Long, (A.) & Sedley (D.). 1987. The Hellenistic Philosophers, Vol. 1: Translations of the Principal Sources with Philosophical Commentary. Cambridge: Cambridge University.


Sextus Empiricus. 1914. Sexti Empirici Opera. Vol. 2, Adversus Dogmaticos, libros quinque (Adv. Mathem. VII-XI) continens, edited by Hermannus Mutschmann. Lipsiae [Leipzig]: Teubneri [Teubner]. Available online:


Online text transcription at:







This entry redoes the one here, made a while ago:





No comments:

Post a Comment