by Umberto Eco
9.9. Duns Scotus and the Modistae
Duns Scotus and the Modistae represent a sort of highly ambiguous fringe between the extensional and intensional positions. In the Modistae we encounter a tortured dialectic between modi significandi and modi essendi. Lambertini (1984) has demonstrated how this point continues for the most part to remain ambiguous, not only in the original texts, but also in the context of modern and contemporary interpretations (see also Marmo 1994).
In the works of Duns Scotus too, we come across contradictory statements. In support of the extensionalist point of view, we find: “verbum autem exterius est signum rei et non intellectionis” (Ordinatio I, 27), while on the other hand, in support of the intensionalist position, we find “significare est alicuius intellectum constituere” (Quaestiones in Perihermeneias II, 541a). There are, however, passages that seem to espouse a compromise solution, opposed to be sure to that of Bacon, according to which, though the thing may be subject to transmutation, this is no reason for the vox that signifies it to change, because the thing is not signified insofar as it exists, but insofar as it is understood to be an intelligible species (Quaestiones in Perihermeneias III, 545 et seq.).
Thus there are scholars who would place Scotus among the extensionalists. Nuchelmans (1973: 196), for instance, referring to the commentary on the sentences (Opus Oxoniense I, 27, 3, 19), declares: “Duns Scotus already affirmed that what is signified by the vocal utterance is a thing rather than a concept.” For others, such as Heidegger (1915),11 Scotus is very close to a phenomenological view of meaning as a mental object. And finally, there are still others, like Boehner (1958: 219), who have no qualms about confessing their ongoing perplexity.12
9.10. Ockham
There has been considerable discussion as to whether Ockham’s extensionalist theory is really as straightforward and explicit as might appear at first sight. If we consider in fact the four meanings of significare proposed by Ockham (Summa logicae I, 33), only the first has an unmistakable extensional sense. Only in this first meaning in fact do the terms lose their ability to signify when the object they stand for does not exist.
That said, even though we cannot be completely certain that Ockham used significari and denotari (invariably in the passive form) exclusively in the extensional sense,13 nevertheless in many passages he did use the two terms with this meaning.
What happens with Ockham—and had already happened with Bacon—is that the semiotic triangle is turned completely on its head once and for all. Words are not connected first and foremost with concepts and then, thanks to our intellectual mediation, to things: they are imposed directly on things and on states of affairs; and, in the same way, concepts too refer directly to things.
At this point, the semiotic triangle would look like Figure 9.7: there is a direct relation between concepts and things, given that concepts are the natural signs that signify things, and there is a direct relation between words and the things on which they impose a name, while he relation between words and concepts is completely neglected (see Boehner 1958: 221 and Tabarroni 1984).
Figure 9.7
Ockham is familiar with the Boethius’s dictum according to which “voces significant conceptus,” but in his opinion it is to be understood in the sense that “voces sunt signa secundario significantia illa quae per passiones animae primario importantur,” where it is clear that illa refers to the things, not the concepts. Words signify the same things signified by concepts, but they do not signify concepts (Summa logicae I, 1). In addition, there exists a rather disconcerting text in which Ockham, taking issue with the notion of an intelligible species, equates it with an image that cannot be more than a sign permitting us to remember something that we have already encountered as an individual entity: what is represented must be in some way already known, otherwise the representing image could never help us recognize the represented object. For instance a statue of Hercules would not help me recognize the real Hercules if I had never seen Hercules before; and I could not tell whether the statue looked like Hercules or not.14
This text assumes (as an issue on which there exists general agreement) the fact that we are not able to imagine, on the basis of an icon, something previously unknown to us. This would seem to be at odds with our actual experience (take, for example, the case of the identikit photo), given the fact that people use paintings or drawings to represent the characteristics of persons, animals, or things they have never seen. Ockham’s position could be interpreted in terms of cultural history as an example of aesthetic relativism: although he lived in the fourteenth century, Ockham was familiar for the most part with Romanesque and Early Gothic iconography, in which statues did not represent individuals in a realistic way, but universal types. When we view the portal of the Moissac cathedral or of Chartres, we have no trouble recognizing the Saint, the Prophet or the Human Being, but certainly not a unique individual. Ockham was unacquainted with the realism of Roman sculpture or the portrait tradition of later centuries.
There is nonetheless an epistemological explanation to account for such an embarrassing affirmation. If the concept is the only sign of individual things, and if its material expression (word or image) is merely a symptom of the inner image, then, without a prior notitia intuitiva of an object, the material expression cannot signify anything at all. Words and images cannot create or implant something in the mind of the addressee (as could occur in Augustine’s semiotics), unless there already exists in that mind the only possible sign of experienced reality, namely, its mental sign. In the absence of that sign, the external expression ends up being the symptom of an empty thought. The subversion of the semiotic triangle, which for Bacon was the end point of a protracted debate, for Ockham is an inescapable point of departure.
There are convincing demonstrations of the fact that Ockham also used significare in the intensional sense—we refer the reader to Boehner (1958) and Marmo (1984) for all of those cases in which propositions maintain their meaning regardless of whether they are true or false. We are not concerned here, however, with Ockham’s semiotics, but with his semiotic terminology. He clearly uses supponere in the extensional sense, inasmuch as suppositio exists “quando terminus stat in propositione pro aliquo” (Summa logicae I, 62). It is, however, equally evident that on more than one occasion Ockham uses significare (in the first meaning of the term) and supponere interchangeably: “aliquid significare, vel supponere vel stare pro aliquo” (Summa logicae I, 4; see also Pinborg 1972: 5).
It is, however, in the context of his discussion of propositions and suppositions that Ockham uses the term denotari. Consider, for example: “terminus supponit pro illo, de quo vel de pronomine demonstrare ipsum, per propositionem denotatur praedicatum praedicari, su supponens sit subiectum” (Summa logicae I, 62). If the term is the subject of a proposition, then the thing the term stands for (supponit) is the one of which the proposition denotes that the predicate is predicated.
In the phrase homo est albus, both terms suppose the same thing, and it is denoted by the whole proposition that it is the case that the same thing is both a man and white: “denotatur in tali propositione, quod illud, pro quo subiectum supponit, sit illud, pro quo praedicatum supponi” (Exp. in Porph. I, 72).
Likewise, denotari is used to indicate what is demonstrated to be the case by the conclusion of a syllogism: “propter quam ita est a parte rei sicut denotatur esse per conclusionem demonstrationis” (Summa logicae III, 2, 23; see also Moody 1935: sect. 6.3).15
The repeated use of the passive form suggests that a proposition does not denote a state of affairs: rather, by means of a proposition a state of affairs is denoted. It is, then, open to discussion whether denotatio is a relationship between a proposition and what is the case, or between a proposition and what is understood to be the case (see Marmo 1984). Through a proposition something is denoted, even if that something supposes nothing (Summa logicae I, 72).
Be that as it may, considering that (i) the suppositio is a extensional category,
and the word “denotation” occurs so frequently in conjunction with the mention of the supposition’ and that (ii) in all probability the proposition does not necessarily denote its truth value, but at least denotes to someone that something is or is not the case,16 we are led to suppose that Ockham’s example may have inspired some later thinkers to use the term denotatio in extensional contexts.
Thanks to the radical shift in meaning of the verb significare between Bacon and Ockham, at this point the term denotare is ready to be considered in an extensional perspective.
It is curious to observe how, if we consider Bacon and Ockham, this terminological revolution first affected the term significatio (involving denotatio almost exclusively as a side effect). But, from the time of Boethius on, the term significatio had found itself so tied in with the concept of “meaning” that it had been able to hold out, so to speak, more courageously against the incursions of the extensionalist point of view. Moreover, in centuries to come, we will continue to encounter significatio, once more used in an intensional sense (see, for example, Locke). Truth-conditional semantics on the other hand was more successful in appropriating the more semantically ambiguous term denotatio.
The cognitivist tradition on the other hand did not follow the lead of using the term “denotation” in relation with meaning.17 Be that as it may, after Mill we find the term “denotation” used more and more in reference to extension.
9.11. After Ockham
Do we have any reason to believe that Mill borrowed the idea of using denotatio as a technical term from Ockham?
There are in fact several reasons to suspect that Mill elaborated his System of Logic with reference to the Ockhamist tradition.
(i) Though he paid considerable attention to the intensional aspects of language, Mill formulated a theory of the denotation of terms in which he makes an affirmation similar to that expressed in Ockham’s theory of supposition: “a name can only be said to stand for, or to be a name of, the things of which it can be predicated” (Mill [1898]1843: II, v).
(ii) Mill borrows from the Schoolmen (and he is the first to admit it—in II, v) the term “connotation” and, in distinguishing between connotative and nonconnotative terms, he states that the latter are defined as “absolute” terms. Gargani (1971: 95) traces back this terminology back to Ockham’s distinction between connotative and absolute terms.
(iii) Mill uses signify in line with the Ockhamist tradition, at least as far as the first meaning assigned to it by the medieval philosopher goes. “A non-connotative term is one which signifies a subject only or an attribute only. A connotative term is one which denotes a subject, and implies an attribute” (II, v). Since the denotative function (in Mill’s scheme of things) is performed in the first place by nonconnotative terms, it is clear that for Mill “to signify” and “to denote” are one and he same. See also: “the name … is said to signify the subjects directly, the attributes indirectly; it denotes the subjects and implies, or involves, or as we shall say henceforth, connotes the attributes … The only names of objects which connote nothing are proper names, and these have, strictly speaking, no signification” (II, v).
(iv) Mill probably opts for “denote” as a more technical and less prejudicial term than “signify,” on account of its etymological opposition to “connote.”
Nevertheless, we have seen that Ockham did not encourage the extensional use of denotare but at most influenced it. Where, in the history of the natural evolution of the term, are we to find the missing link?
The place to look is probably Hobbes’s De corpore I, better known as Computatio sive logica.
It is a matter of general agreement that Hobbes was fundamentally influenced by Ockham, as Mill was by Hobbes. Mill in fact begins his discussion of proper names with an in-depth review of Hobbes’s opinions.
We must, however, note that Hobbes does indeed follow Ockham as far as the theory of universals and propositions is concerned, but at the same time he develops a different theory of signification. For Hobbes in fact there is a clear-cut distinction between signifying (expressing the speaker’s opinions, that is, during an act of communication) and naming (in the classic sense of appellare or supponere, on which see Hungerland and Vick 1981).
Mill ([1898]1843: II, 1) recognizes that for Hobbes names are above all the names of the ideas we have about things, but at the same time he finds in Hobbes proof supportive of the decision that “names, therefore, shall always be spoken of in this work as the names of things themselves,” and the contention that “all names are names of something, real or imaginary … A general name is familiarly defined, a name which is capable of being truly affirmed, in the same sense, of each of an indefinite number of things.” In these passages Mill is close to Hobbes, with the marginal difference that he dubs “general” the names that Hobbes on the other hand dubs “universal.”
However, as we have noted, Mill uses “signify,” not in the Hobbesian sense, but in that of Ockham, and, in place of the notion of “signifying” used by Hobbes, he prefers to employ “connote.” Being deeply interested in connotation, and not realizing that his idea of “connotation” is not all that far away from Hobbes’s “signification,” Mill is convinced that Hobbes privileges nomination (Mill’s “denotation”) over signification (Mill’s “connotation”). In his opinion, Hobbes, like the Nominalists in general, “bestowed little or no attention upon the connotation of words; and sought for their meaning exclusively in what they denote” (II, v).
This decidedly odd reading of Hobbes (as if he were Bertrand Russell) can be explained by the fact that Mill interprets Hobbes as if he were an orthodox follower of Ockham. But, if Mill considers Hobbes an Ockhamist, why does he attribute to him the idea that names denote? Mill claims that Hobbes uses nominare in the place of denotare (II, v), but he had probably observed that, in the De corpore, Hobbes used denotare in four cases at least—five in the English translation of Hobbes’s Latin that Mill probably read, since he cites Hobbes’s work as Computation or Logic.
As for the difference between abstract and concrete names, Hobbes says that “abstractum est quod in re supposita existentem nominis concreti causam denotat, ut ‘esse corpus,’ ‘esse mobile’… et similia … Nomina autem abstracta causam nominis concreti denotant, non ipsam rem” (De corpore I, iii, 3). It should be underscored that for Hobbes abstract names do in fact denote a cause, but this cause is not an entity: it is the criterion that supports the use of an expression (see Gargani 1971: 86 and Hungerland and Vick 1981: 21). Mill reformulates Hobbes’s text as follows: a concrete name is a name that stands for a thing; an abstract name is a name that stands for an attribute of a thing (1843, II, v)—in which “stand for” is Ockham’s “stare pro aliquo.” He adds, furthermore, that his use of words like “concrete” and “abstract” is to be understood as being in keeping with the usage of the Scholastics.
Mill probably extrapolates from this passage in Hobbes the idea that, if abstract names do not denote things, this is instead the case for concrete nouns. For Hobbes in fact “concretum est quod rei alicujus quae existere supponitur nomen est, ideoque quandoque suppositum, quandoque subjectum, graece ypoleimenon appellatur,” and, two lines above, he writes that, in the proposition corpus est mobile, “quandoque rem ipsam cogitamus utroque nomine designatam” (De corpore I, 111,3). Thus, designare makes its appearance in a context in which it is linked on the one hand to the concept of supposition and on the other to that of denotation.
Since concrete names can be proper either to single things or to sets of individuals, we may say that, if there exists a concept of denotation developed by Hobbes, it is still halfway between between Peter of Spain’s suppositio naturalis and his suppositio accidentalis. This is why Hungerland and Vick (1981: 51 et seq.) stress the fact that denotare could not have had for Hobbes the same meaning it has acquired in our contemporary philosophy of language, because it applies, not only to logical proper names, but also to the names of classes and even to nonexisten
t entities. But Mill does buy into such a perspective, and therefore could have interpreted Hobbes’s denotare in an extensionalist mode.
In De corpore I, ii, 7, Hobbes states that “homo quemlibet e multis hominibus, phliosophus quemlibet et multis philosophis denotat propter omnium similitudinem.” Denotation then once more concerns any and every individual who is part of a multitude of single individuals, insofar as homo and philosophus are concrete names of a class. In De corpore I, ii, 7, he adds that words are useful for syllogistic proofs, because, thanks to them, “unumquodque universale singularium rerum conceptus denotat infinitarum.” Words denote concepts, but only of singular things. Mill translates this attitude along clearly extensional lines: “a general name … is capable of being truly affirmed of each of an indefinite number of things” (II, iii).
Finally, in De corpore II, ii, 2, Hobbes writes that the Latin name parabola may denote an allegory (parable) or a geometrical figure (parabola). It is not clear whether denotat here means significat or nominat.
To sum up:
(i) Hobbes uses denotat at least three times in such a way as to encourage an extensional interpretation, in contexts that recall Ockham’s use of significare and supponere.
(ii) Although Hobbes does not use denotare as a technical term, he nonetheless uses it with some regularity and in such a way as to preclude its being interpreted as a synonym of significare, as Hungerland and Vick (1981: 153) have persuasively stressed.
(iii) Hobbes probably moved in this direction under the influence of the ambiguous alternative offered by the passive denotari that he encountered in Ockham, as well as in some logicians belonging to the Nominalist tradition.