Sometime after the publication of the Logical Investigations, Husserl himself abandoned the theory of linguistic meaning as species and introduced a new view of meanings as special abstract entities, which he called noemata.
Anglo-Saxon readers of Husserl – indeed almost all his commentators – have interpreted his development from the Investigations to the Ideas in a way which takes for granted that the teleology which Husserl himself retrospectively inscribed upon his changes of mind – his ‘development’ – has some basis in the facts themselves. Thus it has been assumed that Husserl’s rejection of the theory of meaning as species in favour of the theory of meaning as noema was somehow justified. This assumption seems, however, to be supported only by appeal to what Husserl himself has to say about the matter (after he has already given up the earlier position), and by the presence of a number of peripheral, if intriguing, similarities between the noema theory and the theory of Sinne developed by Frege in his “ber Sinn und Bedeutung”. It is nevertheless interesting to spend some time examining the later theory, particularly in the refined form it has been given by Fllesdal (in a number of papers) and by McIntyre and Woodruff Smith (especially in their book of 1982), since this may be said to combine many of the benefits of the language-based approach with a framework within which mental acts are capable of being taken seriously.16
On Husserl’s earlier (species) theory, if Erna understands what Hans says, then this is because Hans’s and Erna’s thoughts are instances of the same species (at some level of generality), a fact which is itself to be understood in terms of certain kinds of constancy (similarity of parts) in the space of mental acts – constancy which has come about through a certain historical process (Hans and Erna share the same background of habits and skills). On the later (noema) theory no such historical account is possible, for we are dealing not with constancy amidst real variation, but with abstract meaning-entities outside space and time. Hans succeeds in communicating with Erna, on this account, because the meaning of his utterance, a certain abstract noematic Sinn, becomes the meaning of Erna’s act of registering this utterance. It is as if the noematic senses are stars in an abstract heaven to which our successive acts, and even the successive acts of distinct subjects, may – somehow – be identically directed.
So far, so Fregean.
Husserl’s noema theory can be said to be superior to Frege’s in at least one respect however. For Husserl provides an account of the noemata or senses of all acts: perceptual, imaginative, judgmental, etc., where Frege can cope only with judgmental acts, or more specifically with the language bound up with specific sorts of judgmental act. This is an important advantage, since if we want to deal, e.g., with indexical uses of language (‘That bird is flying high’), we shall find it necessary to recognise that the relevant utterances can be meaningful only if they occur as parts of larger wholes which include acts of perception directed toward non-linguistic objects.17 In this very advantage of Husserl’s later theory lies a danger however. For the structure and individuation conditions of abstract noemata are essentially derived from our understanding of the logical structures of corresponding linguistic expressions. (This, surely, is the central message of the new ‘Frege-Husserl semantics’.) If, therefore, we insist that abstract noemata are such as to exhaust the meanings also of perceptual acts, then we would seem to be trying to bring into coincidence two entirely different sorts of structure – having different sorts of multiplicity. For the structures of linguistic meanings are discrete, subject to those sorts of non-continuous variation which come about through the divisions and combinations dictated e.g. by the rules of syntax. The structures of perceptual contents, in contrast, are continuously variable along a number of qualitatively highly specific dimensions, in a way which implies that it is impossible that there could ever be a fitting together of the two of the sort that is required by the newly fashionable interpretations of Husserl’s later theory.
The earlier theory, on the other hand, which draws a sharp line between the two sorts of content, is subject to no such danger. And in other respects, too, the comparison between Husserl’s own successive theories of meaning is not at all to the disadvantage of the former. Indeed neither in Husserl, nor in contemporary proponents of the noema, is any argument given for favouring the later theory rather than the earlier. This is true not least because the earlier theory is not, in any of the secondary literature, worked out in detail.18
When Husserl’s two successive theories are compared, we see a number of advantages of the former. This is so, first of all, in regard to their respective ontological commitments. Both theories accept the need for mental acts as real events (in the later theory these are called ‘noeses’). Further, both accept the need for some account of generality. Indeed the later theory distinguishes (and this is just for starters): the noematic Sinn of an act, the matter of the act which ‘entertains’ this Sinn (a certain real moment), the essence or species of this matter, and in principle also the essence or species of the noematic Sinn.19 The earlier theory, in contrast, has only acts (certain real events), their parts, and the species these instantiate, and we have seen that even the latter need not, of necessity, be taken ontologically seriously. Further, the relations of species to instance and of part to whole to which it appeals are well understood.
The later theory, in contrast, needs relations capable of embracing as their relata both abstract and concrete entities (as if such heterogeneous entities could be capable of being combined together within a single whole). As Woodruff Smith and McIntyre inadvertantly reveal (cf. their diagram on p.134), the intermediary role of abstract noemata saddles us with two insoluble problems:
(i) the problem of entertaining, i.e. the problem of the relation of the noema to concrete mental episodes. (How can a real mental event exist together with an abstract noema/Sinn within a single whole?)
(ii) the problem of anchorage, i.e., the problem of the relation of the noema to concrete objects. (How can a mental act, in somehow grasping an abstract noema, thereby be directed or referred e.g. to a concrete thing?)
Woodruff Smith and McIntyre’s own suggested solution to the first of these two problems operates at the level of metaphor. An act, they say, “intends (is directed toward or is intentionally related to) an object if and only if the act (or its noesis) entertains a certain noematic Sinn and that Sinn prescribes that object” (Woodruff Smith & McIntyre 1982:143). But what is this ‘entertaining’, which appears to be a peculiar sort of non- intentional intentionality?20 It seems, indeed, that the word ‘entertains’ can elucidate nothing. It merely recalls an exactly parallel problem in the interpretation of Fregean philosophy, the problem of giving an account of the relation of Fassen between an act of thinking and a Fregean Gedanke.21 It is perhaps significant that Woodruff Smith and McIntyre (op. cit.) find no further analysis of entertaining in Husserl’s own writings. Rather than accepting this as sufficient evidence of a major flaw, either in their interpretation of Husserl’s theory or in the theory itself – that perhaps a view of the relation between act and meaning along the lines of the earlier theory might be right after all – they take refuge behind a set-theoretical analysis of ‘entertaining’, conceiving it as a ‘many-one or functional relation’ (1982:146). But the rigmarole of functions across possible worlds which they wheel forth in order to support this view (cf. their chs. 6 and 7) yields (at best) nothing more than a rather shaky structural analogy; it does not tell us what the ontological status of these peculiar abstracta and of the associated relations might conceivably be.
In regard to the problem of anchorage, Husserl himself was ultimately to by-pass this problem by abandoning the attempt to establish contact between noema and world (the insidious pressure of the noema theory in the direction of ‘transcendental idealism’). For it seems that the noema theory is unable to specify which objects acts are directed towards. Noemata are abstract entities (they are rather like the ‘concepts’ of old). In no sense are they tied up with or sensitive to the concrete and individual spatio-temporal entities which (as we normally conceive things) people the world of our experience. Thus the noema theorist has no way to distinguish, say, thinking-about- McIntyre from thinking-about-some-philosopher-qualitatively- indistinguishable-from-McIntyre. To succeed in describing a particular person’s thinking-about-McIntyre (his being minded in just this way), we need to recognise that his acts, and their background, are tied to a certain segment of reality, that they are one-sidedly dependent upon certain objects, in the sense of our discussion above (see Smith 1984 for a view along these lines).
Something similar holds even where there is no object of our act, for instance where we are thinking (as we conceive things) about the god Jupiter. For here, too, there is a problem of anchorage: even our acts of thinking-about-Jupiter need to be distinguished from acts of thinking-about-some-god-qualitatively- identical-with-Jupiter, and to this end they must be tied into a certain complex background which includes ancient Mediterranean peoples, their religious beliefs, traditions and practices, and the remnants and reports which survive in various media and inform us of these.22 This is another dimension in which Husserlian phenomenology is inadequate to the ontology of acts: it fails to do justice to this background, which exists in virtue of a network of foundation relations between any given act and the objects of prior acts with which it is associated.
The response of the noema-theorist is to seek to simulate this mundane background by complicating his account of the mutual interrelations among acts, by appealing to what is called the act’s ‘horizon‘ (Ideas I, § 149).
The clearest account of this matter is provided by Woodruff Smith and McIntyre (op. cit.) in their ch. 5. Every act, they tell us, has a horizon, fixed by specific components of the act’s Sinn together with parts of the subject’s conceptual scheme or belief- system. Now it is of course reasonable to seek to extend the phenomenology of the act by recognising a role for background beliefs, even those background beliefs that are not “active” phenomena of consciousness in the way the act itself is (1982:254). We could see this background as somehow latent in an act, present in it ‘in the form of a habitus‘ (Husserl, 1947, § 25, cf. also § 67b). The Husserl of ‘horizons’ and ‘noemata’ is not, however, operating with real events and states and with their real cumulation through time, nor with that background of knowledge that is actually acquired in our past experience. He is operating, rather, with peculiar non-actual dispositions.23 Thus he sees an act’s horizon as consisting of various possible acts, in which the object of the initial act would be intended under various further aspects, with details filled in about (say) those sides of the object that are originally hidden from view (WS & M, p.239). The horizon of an act of perception, for example, would consist of perceptions the perceiver could have had in the past (and indeed in the future also) (Woodruff Smith and McIntyre, 1982:259). The noematic senses of these merely possible components of the horizon are then held, as if by magic, to contribute to determining which object the act is directed towards:
The complete “meaning” of an individuative act, as we have described it, typically includes not only the sense (Sinn) that is actually and “explicitly” present in the act,…but also the system of senses correlated with certain related background beliefs. A complete phenomenological analysis of the act must embrace those senses as well…because it is [they] that ultimately prescribe which individual the act is directed toward or is about. (1982:390)
The problem of objective reference is hereby however shunted off into a corner of dark ‘potentialities’. Not only is this appeal subject to the objection that it is an account of what is real in terms of what is merely possible. Even on its own terms it is not adequate to do the job which it has set itself: no account of the phenomenology of a person’s acts, not even when the implicit or latent ‘horizon’ of these acts is taken into account, not even when their various ‘sedimentations’ are taken into account, can ensure for these acts the appropriate referent. For again, even the complicated horizonal background could in principle occur elsewhere, e.g., on Putnam’s twin-earth. To anchor putative reference to ‘our’ Jupiter and his background as distinct from twin-Jupiter and his background we need an indispensable relational element, and this means breaking out of the phenomenological circle and taking the acts of the subject in their natural setting, intervolved in manifold ways with the rest of the world.
Barry Smith, Husserl, Language, and the Ontology of the Act