Topics in Latin Philosophy from the 12th–14th centuries

Collected Essays of Sten Ebbesen Volume 2

Sten Ebbesen

• Publisher: Routledge
• Year: 2017
• ISBN:
  9781351878777
• Edition: 1
• Topic: Philosophie
• Subtopic: Histoire et théorie de la philosophie

Sten Ebbesen has contributed many works in the field of ancient and medieval philosophy over decades of dedicated research. His crisp and lucid style and his philosophical penetration of often difficult concepts and issues is both clear and intellectually impressive. Ashgate is proud to present this thematically arranged three volume set of his collected essays, each thoroughly revised and updated. Volume Two: Topics in Latin Philosophy from the 12th -14th Centuries explores issues in medieval philosophy from the time nominalists and other schools competed in twelfth-century Paris to the mature scholasticism of Boethius of Dacia, Radulphus Brito and other 'modist' thinkers of the late thirteenth century and, finally, the new nominalism of John Buridan in the fourteenth century.

Chapter 1

Early Supposition Theory (12th–13th centuries)¹

It was one of the medievals’ fundamental beliefs that words in general and nouns in particular are signs, i.e., each word may be compared to a forefinger or an arrow pointing at something else. If you know the language the word belongs to, you will also be able to identify the object pointed at in each case. It was also agreed doctrine that the objects of every-day experience are analysable into forms and the bearers of the forms. The two doctrines were, so to speak, united in the classical definition of a noun, ‘nomen significat substantiam cum qualitate’, ‘a noun signifies something which can be the bearer of a quality (or: property, or: form) as equipped with some quality (or: property, or: form)’.

At first blush, this may seem very simple and OK. If one asks what ‘man’, for instance, signifies, no long meditation is needed before the answer can be produced: ‘A substance, i.e., a self-subsistent entity or bearer of qualities/forms, that is able to sense and move and think and die’ – i.e., the answer consists in stating the definition of man.

Problems begin to arise when the notion of truth enters the picture. Unless interest is limited to ‘scientific’ propositions about non-extensional objects, a correspondence theory of truth requires for verification or falsification of an individual proposition that one can identify the objects signified by the terms of the proposition somewhat more precisely than the above procedure allows one to do.

Another difficulty arises from the substantival use of such adjectives as ‘grammaticus’. There is no natural species of literates, and hence a medieval thinker might find it difficult to accept that ‘grammaticus’ has ‘substance equipped with literacy’ for its significate. On the other hand, it is an undeniable fact that in some propositions ‘grammaticus’ occupies a position that could also have been occupied by ‘homo’ without making the proposition nonsensical, sometimes even without changing its truth-conditions.

To deal with such difficulties Anselm of Canterbury (11th century) introduced a distinction between signifying and being a designation of (significare, appellativum esse, cf. Henry (1974)). ‘Grammaticus’ signifies the quality or form ‘literacy’, but usage allows us to employ it as a designation of a substance, viz. man. In other words he distinguished between meaning (significatio) and reference (appellatio); but to me at least it is not quite clear whether only individuals or both individuals and species specialissimae could be referents in Anselm’s opinion. [36]

Anselm’s specialized use of appellare has no clear model in earlier texts, as far as I am informed. It certainly is not ancient, though one can surely find passages in ancient books that may have provided inspiration for Anselm or whoever invented the terminology. Appellare in the sense of ‘refer to’ stayed in use at least till the late twelfth century; frequently also its synonym nominare was used, and sometimes another synonym, nuncupare. But gradually another term encroached upon the territory of appellare; it was supponere (pro) and by the end of the century suppositio had become the standard word for ‘reference of a substantive noun’, appellare having become more specialized, meaning now ‘refer to objects existing in the present time’.

The theory of supposition with the associated theories of copulatio (sign-capacity of adjectival terms), ampliatio (widening of referential domain) and distributio constitute one of the most original achievements of Western medieval logic. There is nothing really similar in any ancient text the medievals knew – though surely some Stoic writings once contained investigations of the problems these theories deal with – nor had contemporary Byzantium anything similar.

The twelfth century produced a considerable harvest of rules about the referential range of terms in various contexts. When the thirteenth century arrived, a standard terminology had prevailed with such names as suppositio confusa and suppositio determinata for some particularly important types of referential range, and a chapter on supposition had become a standard feature of Introductions to Logic (summulae). But then the development of the theory stopped. It appears that, at least on the Continent, the chapter on supposition in the summulae became one that young students would be taught very early in their career, perhaps before entering university – and then forget all about through the rest of their student career.

In this paper I shall refrain from listing treatises ‘de suppositionibus’; I shall on the whole refrain from following the developments of terminology and systematics. The spade-work in those fields has been done by De Rijk in his Logica Modernorum (1962–67).

I will try to point to and explain some characteristic features of twelfth- and early thirteenth-century speculation about supposition without going into details and without paying much attention to the opinions of individual authors, not even when they protest they disagree with something I say they thought. I am not looking for the particular, but for general attitudes and patterns of thought underlying their investigation of suppositio.

But let me, nevertheless, begin with the origin of the term suppositio. The result of a search for its origins is likely to influence our understanding of early supposition-theory, and previous research into the matter appears to me unsatisfactory. De Rijk, in Logica Modernorum II.1 ch. XVI, has made a point of demonstrating that supponere, suppositum, suppositio are terms of grammatical origin, meaning ‘to put as grammatical subject of a verb’, ‘grammatical subject’ and ‘putting as grammatical subject’ respectively; that no ontological or quasi-ontological sense (‘substratum of accidents’) attaches to them; that the grammarians Priscian (sixth century) and Petrus Helias (mid-twelfth century) and other early [37] twelfth-century grammarians and logicians consistently use the terms in the above ‘grammatical’ sense, the meanings ‘refer, referent, reference’ being secondary, supponere having got the sense ‘be subject’, and hence supponere verbo pro aliquo the sense ‘be the subject of a verb while representing something’. I will gladly admit that the grammarians’ use of supponere is relevant to the history of the term, and that ‘put as subject’ is often the best translation of supponere in twelfth-century texts. But I am not at all happy about the addition ‘grammatical’.

De Rijk starts by considering some passages in Priscian. He claims that in priscian suppositum, a translation of the Greek ὑπΟκ∊ίμ∊νΟν, means ‘subject of a proposition‘, and finds his view confirmed by the occurrence of subicere and subiectio in the same context. The best way to check this interpretation is to look up Priscian’s source, for his Latin grammar is to a large extent an adaptation, very often a direct translation, of Apollonius Dyscolus’ Greek grammar from the second century AD. A major part of Apollonius’ work has perished; thus we have little of the part that corresponded to Priscianus Maior (the morphology). But by good luck we are still in a position to check most of Priscian’s supponere passages in the Greek original.² Reading those passages in Apollonius as well as others in which he uses the word ὑπΟκ∊

σθαι, one realizes that to him it does not mean ‘to be grammatical subject’, not even in the wide sense De Rijk attributes to the medievals (‘both subject-term of a proposition (talk, discussion) and its subject-matter‘) – for even on that broad interpretation, a grammatical subject must qua constituent of a proposition be that constituent which is opposed to the predicate and which, if expressed as a noun or a pronoun, must agree with the predicator-verb in number and person, and be in the nominative case. Apollonius’ use of ὑπΟκ∊

σθαι ‘be underlying’ is different, and nothing suggests that Priscian misunderstood him.

To Apollonius and Priscian, that which is ‘underlying’ is what is within the range of a name, that which the name refers to or which is talked about, that which is the bearer of a quality etc.; but not the subject as constituent of a proposition as opposed to the predicate. To take but one characteristic example of the use of the word, one version – and no doubt the original Stoic one – of the definition of the noun runs³ ὄνΟμά ἐστι μέρΟζ πτωτικόν, ἑκάστῳ τ

ν ὑπΟκ∊ιμένων σωμάτων ἢ πραγμάτων κΟινὴν ἢ ἰδίαν πΟιότητα ἀπΟνέμων ‘A noun (name) is a case-marked constituent of discourse that attributes a common or an individual quality to each of the underlying bodies or activities’; in Priscian’s Latin translation the passage runs⁴ ‘Nomen est pars orationis quae unicuique subiectorum corporum seu rerum communem vel propriam qualitatem distribuit’. The observation that in ancient grammar ὑπΟκ∊ίμ∊νΟν does not mean ‘subject’ [38] is not new. The Byzantine grammarian Choeroboscus⁵ expressly mentions that grammarians have a peculiar use of the word. The subiectiones in one of the Priscian passages need not worry us: they are not ‘subject-puttings’ but ‘substitutions’, subiectiones nominativae rendering ἀνθυπαγωγαὶ ὀνΟματικαί.

If we turn to Petrus Helias, we find that to him, for something to be suppositum is to be the bearer of a name or of a form/quality, exactly as in Priscian and Apollonius. But as an innovation (comparing with the ancients) he also on several occasions contrasts suppositum and appositum, the subject and the predicate of a proposition. It has been said that Petrus Helias uses the word ‘suppositum’ in two ways, but it is not so clear that the two senses would appear different to him; he and many contemporaries who speak in the same way, might say that ‘Well, what’s the difference? In either case by suppositum I mean that which is the bearer of a name or a form; when I contrast suppositum and appositum, I mean by the former that which is the bearer of the form attributed to it by the appositum’.⁶

It is this notion of ‘bearer of form’ that provides the link between the divers medieval theories of supposition and explains some features of these theories that have been considered weird. It is a natural thing, once you have begun to ponder about reference, to ask, ‘What or which things are the bearers of the name ‘man’ and of the form of man quite generally, without any respect to a context in which the word is used?’ And it is natural to answer, ‘All men, present, past and future alike’. But however natural the question may be, it is of no great importance to anyone to know the answer. Finding the referents of a word is only interesting when that word has been used in a context.

It is also natural to ask ‘What or which things are the bearers of the name ‘man’ and the form of man that we refer to when we say “yellow man”?’ And it is reasonable to answer that we are here referring to a sub-set of man, only yellow men being taken into consideration. But again, the answer is of no real interest if ‘yellow man’ is thought to occur out of propositional context.

It is immensely more interesting to know which are the referents of ‘man’ and ‘yellow man’ if we have a propositional context. Imagine some psychologist claiming that if we disregard the marginal cases at the left side of a bell-shaped curve, it is a fact that all yellow men have an IQ above 100. A fellow-researcher would like to check the claim. To do so, he must know exactly who are the referents in the original claim. An elimination procedure can then be applied: They are men; they are not all men, only yellow men; they are not all yellow men, but only presently existing yellow men. The elimination of past and future men may be problematical, but it is, of course, of crucial importance to know whether the claim to be examined regards yellow men of all times or only men of some definite time. When the elimination procedure has arrived at its end, one can start empirical research with a view to verifying or falsifying the claim.

Medieval logicians of the twelfth–thirteenth centuries did not in general subscribe to some strong variant of the identity theory of predication, and they tended not to [39] occupy themselves with sentences that contain a transitive verb, like ‘Socrates beats Plato’; hence they would naturally concentrate their investigation of reference round the subject terms of sentences. And so they did. Some went so far as to claim that the question of supposition arises only with regard to subject terms; some less restrictively said that the question arises only with regard to terms occurring in a proposition; some, finally, would admit of supposition also for a word or phrase considered with no respect to context. To some thinkers, the choice of one alternative to the exclusion of the two others may have been a matter of conviction and concern for coherence in their logic as a whole. But I doubt if this was always the case, and I think one does well not to overstress the differences, in particular because our principal sources for the theory of supposition are all treatises meant for use in elementary instruction, often not very carefully composed, and often, we may presume, the work of second-rate masters and/or students’ compilations of the master’s teaching.

I suppose that if we were to ask a twelfth–thirteenth-century schoolman about the ultimate use of knowing about supposition, he would say that without such knowledge we should not be able to judge the truth value of such important propositions as ‘Christ was a ...