Epistemic Modality, Language, and Conceptualization: A Cognitive-Pragmatic Perspective

Human Cognitive Processing

Volumn 5, Issue , 2001, Pages 1-450

  Nuyts, Jan ^a  

^a UNIVERSITY OF ANTWERP   (Belgium)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 23844533343     PISSN: 13876724     EISSN: None     Source Type: Book Series    
DOI: None     Document Type: Review

References (474)

1. Harder (1996: 56ff) takes my use of the notion of a 'cognitive infrastructure' (with specific reference to Nuyts 1992b) to imply that I hold a 'pancognitivist' view, i.e. a view in which there is no room for an external world and everything is considered to be inside the mind. This is a misunderstanding, however. I do not deny the existence or the relevance of an external world at all, on the contrary. And my use of the notion 'infrastructure' is intended to be completely theory-neutral, i.e. similarly as when one is referring to the infrastructure of an institution like a university, e.g. In the discussion, Harder refers to my view that social structure is entirely based on cognitive structure, and does not have an existence independent of the cognitive systems of the individuals constituting the social group. (Which also implies that cognition is a thoroughly social phenomenon.) This is a correct rendering of my position, and I fully stand to it. But this view in no way implies that there is no outer world, including other individuals with whom the cognitive system holds a social relationship (cf. the fact that I am referring, not to one single cognitive system, but to a set of cognitive systems constituting a group). For example, when I refer to the speaker's interlocutor (or to the hearer), I am not referring to some mystical apparition or a fata morgana in the speaker's mind, but to a real individual in flesh and blood out there in the world, as perceived and interpreted by the speaker.
2. Language also has non-communicative uses, of course, but these may be considered secondary and derivative. See Grice (1969: 174ff), Searle (1972), Nuyts (1993b).
3. Thus, 'pragmatic' here refers to a (functional) perspective on language, not to a component of language besides syntax and semantics. The concept of a pragmatic component does not figure at all in the present framework. See Verschueren (1985), Nuyts (1992b: 64-71), Verschueren et al. (eds. 1995).
4. 'Cognitive Linguistics' will henceforth refer to the (originally Californian, though by now quite international) research tradition à la Talmy (1988a, 1988b), Langacker (1987) and Lakoff (1987). Non-capitalized 'cognitive linguistics' will stand for any type of linguistics concerned with the cognition of language, including the one represented in this monograph.
5. These are trends, of course, and there are notable exceptions to them: e.g., Bruner (1990) and Clark (1992, 1996) on the cognitive side, or Givón (1990, 1995, 1998) and Chafe (1994) on the functionalist side, to name just a few.
6. The present notion of functionalism, common in linguistics and most of language psychology, should obviously not be confused with the notion in cognitive psychology, involving the view that cognition can be studied independently of the neurophysiology of the brain (e.g., Fodor 1968; Johnson-Laird 1983).
7. In essence I also subscribe to the latter view (see Nuyts 1992b: 7), and most cognitive research in linguistics and language psychology is in actual practice functionalist in this sense. But this matter is immaterial for the present purposes.
8. Many may find this disputable for the category of emotions, in view of the in western thinking deep rooted division between the domains of rationality and emotionality, and the tendency to associate notions such as 'conceptual' and 'cognitive' with the rational domain only. This division may be detrimental to our understanding of these phenomena, however. In fact, the borderline between the domains of emotions and of intentions and attitudes is far from sharp (cf., e.g., deontic or volitional attitudes).
9. Dik has speculated that a representation of type (1b) might also be adequate to model conceptualization (see below). He did not have the time to systematically introduce this in Functional Grammar, however, Dik (1990) being a rare (if not the only) attempt to draw consequences for linguistic analysis. Otherwise, his suggestions have had little or no effect on research in this framework.
10. The narrow focus even in current research clearly appears from the state of the art overviews on sentence parsing in Mitchell (1994), or on word recognition and lexical access in Balota (1994), Lively et al. (1994) and Simpson (1994), or on language production in Levelt (1989) and Bock and Levelt (1994).
11. In AI models the 'knowledge database' is often even further developed than the language processor, but the problem with much of AI is that it aims to develop user-friendly systems for different practical purposes (translation, data-storage, interfaces, etc.), rather than a cognitively plausible model of human language processing or conceptualization.
12. Whorfians are obviously likely to adopt propositionalism: cf. Whorf's (1956: 252) own view that “thinking itself is in a language - in English, in Sanskrit, in Chinese”. Yet most of the developers of propositionalist models implicitly or even explicitly (e.g., Jackendoff) adopt a non-relativist, universalist view of conceptualization. Dik is one of the rare exceptions.
13. Significant for the state of the art in conceptualization research is that in this imagery debate, too, much of the controversy has been over the question what counts as valid evidence for or against imagery as a representational format in conceptualization.
14. Some - e.g., radical connectionists - believe that none of this is really an issue, because all there is is hardwired neural nets. I will not be concerned with this line of reasoning here, as I do not think a radically materialist view of cognition is maintainable or fruitful (see Nuyts 1992b: 7ff). This is irrespective of the question whether connectionist nets are plausible as a model of the neural systems related to the cognitive domains of language and thought.
15. Hence, ideally, investigation of conceptualization should be multimodal. Unfortunately, such research is still very scarce. Promising in this respect is the recent upsurge of studies on gesture and how it combines with language (e.g., McNeill 1992, 1997; Kita 1993).
16. I, for one, do believe that cognitive systems as described in our theoretical models (if they are adequate) are somehow coded in the neurological system. I would not use the term 'stored', either, however. And like Langacker, I do not believe that we can 'see' them by looking at the brain (see Nuyts 1992b: 6ff). But this dispute is immaterial for the question whether one should accept a procedural concept of grammar, pace Langacker's suggestion.
17. Despite the differences in the theories, Langacker's argument is akin to the mystification found in Chomsky's discourse about what a Generative Grammar is or does. Chomsky claims that his model does not produce anything but relates meanings and phonetic patterns in a neutral way. Yet Generative Grammar (in whatever version) formally is a productive system which does not neutrally relate meanings and forms at all (see Nuyts 1992b: 125-126). Langacker's argument has in common with this that there is a disconnection between what the model is claimed to stand for, and what the formalism of the model actually renders.
18. All of this is completely irrespective of the question whether connectionist type models are adequate to understand the neurological part of the business.
19. The notion of 'truth' as used in this definition should not be interpreted as a theoretical notion (as it occurs in truth-conditional semantics). Rather, it refers to a general pre-theoretical concept people have and somehow use in thinking about the world, which is thus a subject of investigation. Whenever I use the notion, it is in this sense.
20. Especially in formal semantic approaches, one sometimes discerns a tendency to deal with epistemic modality (and modality in general), not in terms of a scale, but in terms of discrete categories, usually (only) possibility and necessity. From a cognitive and functional perspective, however, there are strong arguments to assume that humans do think in terms of a scale, even if for most purposes they will probably only distinguish rough positions on it. Thus, English, Dutch or German have basic terminology only for, most precisely, certainty, probability, possibility, improbability, and impossibility. But speakers can further scalarize these positions by means of grading expressions (very probable, rather certain, not entirely impossible, etc.). And ultimately, if really needed, they can even quantify likelihood (e.g., a 90 percent chance). It is hard to see how a 'discrete categories' approach can handle these facts.
21. Some have related the epistemic scale to concepts such as 'force dynamics' (Talmy 1988a; Sweetser 1990) or 'degrees of distance' (Fleischman 1989). No question that the cognitive concept of degrees of likelihood may be related to such other conceptual categories, e.g., in that expressions of temporal distance get metaphorically used to express degrees of likelihood, as Fleischman shows. But any suggestion that the former might be reducible to the latter is unwarranted. Veridicality of information is - intuitively and observationally - too basic a category in a human's interaction with the world (see below) to be explained away.
22. The term 'modality' is sometimes also used to cover any kind of speaker qualification of a state of affairs (i.e. including emotional attitude, evidentiality, or even tense and aspect). This use is most common in philosophy (see Perkins 1983: 6ff and Palmer 1986: 9ff for references), but occasionally also occurs in linguistics (e.g., Ransom 1977, 1986; Dietrich 1992). So used, the term is synonymous with the present notion of a qualification.
23. Hence, also Bybee's (1985) and Bybee et al.'s (1994: 177-180) distinction between epistemic and speaker-oriented modality (forms such as imperative, optative, permissive, etc.) is terminologically unfortunate, as epistemic modality is extremely speaker-oriented.
24. This paradigmatic approach is obviously applicable to any other semantic or functional domain in language, with the same potential for investigating the research question formulated in Section 1.2. Unfortunately, it is quite rare in empirical research in linguistics in general.
25. I am not aware of any better or more explicit characterization of the distinction elsewhere in the literature. In most cases the distinction is simply introduced by appealing to its intuitive plausibility, and without any further attempt to explicitly define it.
26. Sanders and Spooren (1996) relate subjectivity with non-observable, knowledge-based evidence and non-subjectivity with observable evidence. In the present analysis this is not maintainable. Direct observation is more likely to offer good quality evidence than inferencing, but there is no necessary link. Directly observed evidence can be poor quality: e.g., getting a glimpse of something. And inferential evidence can be excellent, e.g. if the premises are based on good direct observation and the inference has been strictly syllogistic. And there is no apparent link at all between observable vs. knowledge-based evidence and the question whether this evidence is known to the speaker alone or is shared by more people.
27. See the papers in Stein and Wright (eds. 1995) on the pervasiveness of subjectivity in language, even if the notion is clearly not always used in the same way (in these papers, or elsewhere; see also below).
28. I see no typological correlates for the definition in terms of the (good vs. poor) quality of the evidence, however. This further adds to the impression that Lyons' original definition of the dimension of subjectivity is not really to the point. See Nuyts (2000b) for further discussion.
29. DeLancey (1997) argues that mirativity constitutes a system distinct from, even if intimately related to, real evidentiality (the marking of types of information sources). This means splitting up evidentiality as a cover notion for everything to do with sources of information into finer categories. I will not take issue with this here. For the present purpose I will stick to the global notion of evidentiality and handle (inter)subjectivity as a subtype of it. See also Section 6.6.
30. Again, Traugott's notion is somehow akin to Langacker's, but the overlap is far from complete (cf. Traugott 1995) and the precise relation is far from obvious. But see Carey (1995) for an attempt to explicate the relation.
31. This might be a reason to avoid the term 'performativity' here, but the problem is that there are no good alternatives. 'Mention' (for 'descriptivity') versus 'use' (for 'performativity') might seem to do the job. But this terminology has been used for yet another purpose which does not conform to the present (cf. Sperber and Wilson 1981). And it leads to formulations which are confusing (cf. the often unwanted ambiguity in statements such as 'the use of the modal adverb') or unattractive (cf. the stylistic oddness of 'the use use of the modal adverb'). On the other hand, using the same term does have the advantage that it signals the fact that the basic mechanism underlying these phenomena is probably identical in both (see Section 6.6).
32. There may be a few circumstances in which the scope of the epistemic qualification could influence the choice of an expression to some extent. Specifically, when the qualification does not affect the main body of the utterance but only a secondary (often parenthesis-like) element in it, if that element does not contain a predicate, use of a modal auxiliary appears impossible for formal reasons. Such cases are marginal, however.
33. As should be obvious from the foregoing, I do not share the attitude not uncommon in corpus linguistics that all there is to good research is a careful study of the data, which will automatically generate the theory (if the latter is needed at all). Such a radically empiricist attitude (it comes in degrees, of course) is unfruitful, as it does not allow the formulation of critical research issues such as in Section 1.2. And it is not maintainable, for any data analysis unavoidably implies a categorization of data, which necessarily goes beyond pure empiricism. The belief that analytical categories emerge from the data all by themselves is a dangerous illusion: data do not speak, they can only be interpreted. Even radical empiricists do introduce categories (often unwittingly), but these run the risk of being ad hoc because they are not motivated by a global perspective (a theory) on the subject matter. As the present investigation will hopefully demonstrate, good science requires combination of a strong commitment to data with a clear theoretical view of the matter.
34. Full reliance on constructed examples is, of course, a risky method for investigating any type of linguistic phenomenon, even a purely syntactic one. See Chafe (1994), Nuyts (1992b: 120ff, 1996a).
35. Of course, it is an illusion to think that one can ever do without intuitions in analyzing corpus data anyway: even for actually occurring cases, intuitions are crucial to understand what they stand for, and what they may reveal about the issue under investigation.
36. Jointly with Wietske Vonk, I am currently further elaborating the experimental testing of the findings of the present corpus-based investigation.
37. Uit den Boogaart (1975: 25) himself observes that these subcorpora are generally very similar as compared to the categories of literary prose and spoken language, in terms of the quantity and quality of the lexical items used and their grammatical features.
38. These adverbs and adjectives can also be used at the term level: cf. John is a probably too weak candidate, John is a probable candidate. The Dutch and German corpus data indicate that so used these forms often have no epistemic meaning anymore, and Chapter 5 will show that epistemic uses of term adverbs and adjectives in Dutch are very rare. So I will limit my attention to sentence adverbs and predicative adjectives here. Hence, when I use the notions 'adverb' and 'adjective' below, I mean 'sentence adverb' and 'predicative adjective', unless indicated otherwise.
39. Unlike English, Dutch and German normally do not morphologically mark the difference between the adverbial and the adjectival form. However, for this specific lexical item there is a tendency in Dutch and a near obligation in German to mark the adverbial form, viz. mogelijkerwijs, möglicherweise.
40. According to the standard dictionary of Dutch (Van Dale Groot Woordenboek der Nederlandse Taal), wellicht is equivalent to misschien. Yet, at least in Southern Dutch, in actual use it is clearly stronger than the latter.
41. On how to interpret an asterisk in constructed examples, see Section 1.6. The utterances in (13) are impossible for information-structural reasons, as I will argue later in this chapter. Still, during presentations I have repeatedly been confronted with the remark that syntactically they do seem possible. Surely, the same pattern with another adverb, e.g. did they often run out of fuel?, is fine. This issue is futile, however. These expressions simply do not and cannot occur for deeper functional reasons, and that makes the question at the syntactic level superfluous. They are simply never constructed, so what would it mean to wonder whether they are syntactically possible or not?
42. That performativity and sub/objectivity should get confused is not too surprising. Performative uses of modal expressions are in a way always subjective if taken in the sense of 'involving a personal commitment of the speaker'. Descriptive uses, on the other hand, are always objective if taken in the sense that they do not indicate the speaker's personal view, but originate outside his/her present circumstances. However, it is essential to see that (inter)subjectivity as defined here (as well as the classical sub/objectivity distinction for that matter) and commitment are two separate matters, as should be clear from the exposition of the two factors in Section 1.4. This is not to say that these two factors cannot occasionally interact, however: see, e.g., Section 2.3 below.
43. The cases I have encountered are overwhelmingly predicative adjectives. Of those less than 10% is epistemic or, most often, ambiguous between a dynamic and an epistemic reading. This is quite comparable to the Dutch modal auxiliary kunnen 'can/may', which will be analyzed in Chapter 4. The few adverbial cases seem in general clearly epistemic, however.
44. For the sake of simplicity, I will henceforth reframe the positions in the literature regarding the old sub/objectivity distinction in terms of the alternative analysis of (inter)subjectivity assumed here, thereby disregarding possible differences between the old analyses and the present one.
45. The effect of types of negative marking in the main clause of propositional attitude expressions on the degree of (inter)subjectivity of the expression is discussed in Nuyts (1990a: 569-570, 1992b: 295-296).
46. This is not true for all adverbs: manner adverbs, e.g., as in he painted the wall quickly, are covered by the past tense of the predicate. This difference is due to the different semantic status of these adverbs. How one should understand this will become obvious in Chapter 6. The relevant point for the present discussion is, then, that the adverb's interpretation is not already determined by the syntax of the utterance (unlike for the adjectives), but is a purely semantic matter.
47. So this is not a case in which a performative epistemic qualification is within the scope of a temporal one, a situation which should not occur since (as will be discussed in Chapter 6) epistemic modality has wider scope - i.e. is higher in the hierarchy of qualifications - than temporality. The epistemic evaluation is within the scope of the irrealis expressed by the past tense form though: it is situated within the possible world evoked by the author.
48. These examples from Lang have been slightly adjusted for the present purpose: Lang uses the emotional attitude adjective bedauerlich 'unfortunate' and adverb bedauerlicherweise/leider 'unfortunately', for which the same argumentation applies. This, again, goes to show that the element of performativity is not only relevant for an analysis of epistemic modality.
49. The findings in Tiggeler (1991) suggest that this is again a property, not only of the epistemic adjectives, but of the category of evaluative predicative adjectives in general.
50. Some functionalists will consider the assumption that one semantic unit should often get expressed in two lexical forms unacceptable, due to their rejection of abstractness in linguistic analysis (see Section 1.2). Surely, so they argue, if there are often two forms at the surface, this must be because there are two underlying forms as well. Yet the present explanation for the frequent occurrence of two surface forms for one underlying unit is thoroughly functional, but it also grasps a basic intuition about the conceptual nature of the phenomenon. This goes to illustrate the point made in Section 1.2 that cognitive functionalism is perfectly compatible with, and even requires some kinds of abstractness in one's analyses.
51. There is one case in the German spoken data which might possibly involve embedded clause negation. Yet its interpretation is highly uncertain because the expression is grammatically infelicitous and while producing it the speaker constantly hesitates, restarts and repairs, thereby conflating different syntactic patterns. Therefore the case has been excluded from the analysis.
52. This is true for the default reading of this expression type, in which not improbable is one functional unit. There is also another reading in which not negates improbable (see (l) below). Such a reading only seems possible in echo utterances, to deny an explicit statement regarding the improbability of the state of affairs in the foregoing utterance. The assumption that the default reading of this expression type does not involve negation of a negative modal qualification, however, is also supported by the fact that it does not carry the presupposition that the opposite could have been true (utterances with 'normal' negation always do carry this presupposition). Thus, in example (45) there is no presupposition that it could have been 'improbable that I (i.e. Luns) will become secretary-general of it (i.e. NATO)'.
53. One might get the impression that in some of these examples - especially in (54) - the modal adverb has the negative expression in its scope. This would go against our assumption that normally polarity and epistemic modality are two dimensions of one complex qualification, coextensive in scope. But this impression is probably only partly correct. An expression such as nooit 'never' is a surface-linguistic contraction of niet ooit 'not ever', in which the negative element has scope over the temporal one. Hence semantically the modal adverb probably only has scope over the temporal component of the expression, not over the negative component.
54. I have not included complement clauses in this, because they normally do not belong to the category of embedded clauses which can be said to provide background information. Cf., e.g., the complements of communication predicates such as he writes that … in (56) above, or of mental state predicates such as I think that …, to be discussed in the next chapter.
55. 'Background vs. foreground' is thus not to be confused with 'old vs. new' (or 'presupposed vs. asserted'): in principle, background information is/can be newly asserted information in a clause (cf. Givón 1984).
56. This indicates that the notion of a 'sentence adverb' should be used with care. Adverbs of this type are clearly not term adverbs, since they are not part of the internal organization of a term, but belong to a higher level of organization. Thus, they are of the same category as the real sentence adverbs to the extent that they operate somewhere at the level of sentence structure, but they nevertheless do not affect the entire sentence (or even the asserted part of it - see Section 1.4c).
57. ?-?certainly I am coming home tonight are all rather marginal, while their equivalents with sentence-internal position of the adverb are all fine. Another strong adverb such as surely in English (which is probably more evidential than epistemic though) does allow preposing, however. I have no explanation for this observation.
58. Excluded from the relevant cases are, for Dutch and German, the (respectively) 48 and 23 backgrounding cases mentioned in (g) above, plus (respectively) 11 and 10 cases in which the adverb occurs in a complement clause. The adverbs in sentence-initial position do not include the few absolutive cases mentioned in (f) above.
59. By 'considerative' I mean that the speaker indicates that (s)he is taking up a suggestion uttered by the interaction partner which, after having thought about it, (s)he considers an acceptable proposal - acceptable to the extent indicated by the epistemic adverb. These Dutch and German particles are actually not completely equivalent in the nuances they express.
60. Believe's use as a prime example in the literature, then, is not due to prototypicality (though believe may be relatively more important in British English), but to the fact that this predicate (and its nominal counterpart) figures as a crucial philosophical notion.
61. Aijmer (1997: 16ff) claims that all mental state predicates, like other epistemic expressions, can be plotted on a scale of 'degree of reliability'. The status of this notion is not entirely obvious to me, but if it should mean 'degrees of strength of the epistemic judgment', I disagree.
62. The former group of predicates is thus comparable to adverbs such as presumably or supposedly, which also have an inferential meaning possibly in combination with an epistemic one (see Section 2.1). It is no accident that these adverbials are all derived from corresponding mental state predicates, of course. For think or believe, however, there are no real equivalents among the adverbs and adjectives. As mentioned in Section 2.2, the adjectives' (inter)subjective meanings are syntactic, not lexical; as I will argue in Section 3.2 below, for the mental state predicates these are lexical.
63. Aijmer (1997: 12ff) suggests that one can distinguish between 'belief' and 'opinion' uses of think. I am not sure how to understand these notions, but if they stand for epistemic and evidential uses in the present analysis, I do not subscribe to her view. I do agree, intuitively, that some qualificational uses of this predicate may be more evidential and others more epistemic, but that does not mean that in such cases the other meaning is not present as well. In any case, in analyzing corpus data (for Dutch denken, at least) it turns out impossible to separate these types of uses: even if one meaning sometimes appears to dominate, the other can never be excluded.
64. Hengeveld (1988) calls these readings 'de re' and 'de dicto', respectively. On earlier occasions (Nuyts 1992a) I have used these terms, too, but I will not further do so since they do not adequately render the difference between the two readings in the present analysis. The non-qualificational reading does not belong in the domain of reported speech. Moreover, these terms are used variously in the philosophical literature, and are therefore potentially confusing. Particularly concerning propositional attitudes, they are sometimes used more or less like Hengeveld's notions (see, e.g., White 1975), but most often they are used to characterize a completely different distinction between readings of the predicates, viz. the one discussed above concerning belief-ascriptions such as 'Ralph believes that the man with the brown hat/Ortcutt is a spy'.
65. See also Vet (1994: 66), although I disagree with his analysis to the extent that he does not seem to distinguish between descriptive qualificational uses and non-qualificational uses of mental state predicates (see later in this chapter).
66. The meaning of these predicates in (82)-(83) vs. (80)-(81) is not always fully identical (e.g., (82a) appears a case of what Aijmer 1997: 12 calls the 'intentional' meaning of think), but is clearly related. The nature of the differences is beyond the scope of my present concerns.
67. Thompson and Mulac (1991) suggest that in English the qualificational reading of these predicates in the complementing pattern only occurs when the complementizer is omitted. This is clearly not true. For think in particular, even the complementing pattern with an explicit complementizer can practically only be understood qualificationally (see below).
68. Only know and doubt and their Dutch and German cognates appear to be exceptional in not allowing a parenthetical usage. One does have uses such as it's difficult, I know, but let's try it anyway. This does not involve the inferential/epistemic meaning of know mentioned above, however. Rather, it involves an argumentative, concessive meaning, which is directly based on the non-qualificational mental state meaning. For some reason the qualificational meaning 'it must be so' seems impossible in parenthetical uses of know. Why there should be no parenthetical doubt will be discussed in Section 3.4 below.
69. Urmson (1952) uses the term 'parenthetical' quite differently, covering all instances of mental state, communication, perception, and other predicates in their qualificational (as opposed to non-qualificational) usage, irrespective of their syntax. That is, they are also called such when they are syntactically complementing. His quite idiosyncratic usage of this term is thus semantically motivated, while I am using the term to characterize a syntactic property.
70. Pace my own earlier assumptions (Nuyts 1992a). The problem is that I failed to distinguish between the non-qualificational and the descriptive epistemic reading (see Section 3.3) of these predicates. What I then called the 'de dicto' reading (following Hengeveld's 1988 terminology) is actually sometimes one, sometimes the other reading.
71. One may wonder why the non-qualificational use of these predicates is so variable in allowing a that-complement. After all, as mentioned above, even the mental state or process reading of these predicates involves a state of affairs. And on a naive interpretation, the function of a that-complement would precisely be to allow the expression of that state of affairs. One line of thought might be to assume a richer meaning for a that-complement than this naive one. Frajzyngier (1991, 1995) has argued that complementizers are independent markers of modality (in a wide sense), and that that in particular is essentially a marker of hypotheticality. This meaning would have been derived from the original usage of the complementizer in the context of verbs of saying, where it served, and still serves, to introduce a 'de dicto' domain: reported speech is obviously less certain than reality itself.
72. Thus, according to Frajzyngier the difference between the non-qualificational direct perception reading and the qualificational indirect inference reading of a predicate such as see (see above) is due to the hypotheticality meaning of the complementizer (which is present in the latter but not in the former reading). In this view, it is not surprising that the qualificational mental state predicates allow omission of the complementizer, since it is redundant. Still, this cannot explain why non-qualificational suppose or know do, but non-qualificational think and to some extent believe do not favor a that-complement: e.g., what is 'known' can hardly be considered more hypothetical than what is 'thought' or 'believed'. (And I will later suggest a completely different explanation for complementizer omission in the qualificational uses of these predicates.) Clearly, there must be more going on, but investigating this is beyond my present concerns.
73. As mentioned, German denken is hardly used qualificationally. A sample of 60, 000 words of spoken and written data each reveals 31 spoken and 19 written cases, of which only 2, both spoken, are qualificational (i.e. 6.5% of the spoken cases, or 0.33 per 10, 000 spoken words). Dutch geloven, however, is often used qualificationally, at least in spoken language, yet still considerably less than denken (cf. Table 3): the spoken VDB-corpus reveals 60 cases of which 41 are qualificational (i.e. 68.3%, or 4.47 per 10, 000 words); the expository VDB-corpus shows 4 cases of which 1 is qualificational (i.e. 25%, or 0.19 per 10, 000 words).
74. Not surprisingly, then, in literary prose (not further considered in this chapter), which is also to a considerable extent narrative, there is a very high frequency - 9.16 0/0000 - of non-qualificational denken as well (again to a considerable extent quotes).
75. Since for the mental state predicates I only consider data from expository prose and spoken language, I have omitted the frequencies of the adverbs and adjectives in literary prose in Table 4. The global frequencies for these expressions have been recalculated correspondingly.
76. Some Dutch speakers (albeit a minority) use geloven instead of denken as the default qualificational mental state predicate (possibly due to dialectal background). The purely subjective cases of geloven in the data may be due to the speech of such speakers.
77. The parenthetical in (105) was only added after a short pause by the speaker, as if he first had to detect some hesitation to answer the question on the part of the interlocutor. So it was probably not planned in advance, which may help explain its mild effect here.
78. Using a complementing pattern instead of a parenthetical in (110) would trigger yet another meaning, not discussed in Section 2.3, viz. that Peter drinks another schnaps because the speaker believes he is addicted. But this issue is immaterial for the present discussion.
79. 7 Dutch and 10 German cases involve an infinitival complement instead of a that-complement, as in he thinks to have won (difficult in English but fine in Dutch and German). This may be considered a reduced variant of the pattern with a that-complement, though, which is preferred when there is referential identity between the main and embedded clause subjects. See also Noël (1997), who shows that contextual factors such as recency of the referent rather than semantic factors determine the choice for either complement in English believe.
80. For Dutch geloven the share of parenthetical forms is much higher: the sample mentioned above contains 19 instances, i.e. 45.2%. Why this is so is unclear to me.
81. Some English speakers appear to accept a negative parenthetical in the context of a negative main clause (but only there), as in it is not dangerous, I don't think, to run out of fuel in the desert. In such a case, the parenthetical merely echoes the negation in the main clause, hence it is not a counterexample to our present point either: the parenthetical is not in focus. Such cases are entirely out of the question in Dutch and German, however.
82. These findings (including those for the parentheticals) clearly run counter to Tottie's (1983) observation that mental state verbs in English show a high tendency (68% in written, 53% in spoken corpus data) to draw negation. On the other hand, the experimental data to be reported in Chapter 5 are more in line with Tottie's observation. In the present case, the difference between the corpus findings and the experimental data is obviously due to the special nature of the experimental conditions. So the question is to what extent Tottie's findings are not also due to accidental special characteristics of his corpus selection.
83. Apart from these, there is one Dutch negative case in a 'reduced' question, viz. je denkt het niet? 'you don't think so?', which echoes an identical statement by the interlocutor, and is thus fully in line with the observations about the reduced cases to be discussed in (b).
84. French has a comparable pattern with an obvious complementizer: je pense que oui/non.
85. There is also the negative mental state predicate betwijfelen, bezweifeln 'doubt', of course, which may be compared to the negative-incorporated adjectives, but which is not covered by the present corpus analysis. A quick count in the same corpus as used for denken and glauben reveals only 3 Dutch and 8 German qualificational cases of this predicate, however.
86. The relation between main and embedded clause negative variants here is thus clearly different from that in the adjectives (see Section 2.4a). So the phenomenon traditionally called 'negative raising' (see Nuyts 1990a, 1992b) is less uniform than has sometimes been assumed (including by myself in the references mentioned).
87. There are also several questions with a reduced pattern in the data, i.e. of the type what do you think?, but I am not concerned with those here. On questions, see (g) below.
88. One Dutch case stands somewhat apart, viz. al wordt dat door de buitenwereld wel eens gedacht 'even though this is sometimes thought by outsiders'. It is closest to (121a), but then used as a metaremark to the speaker's own statement. I have nevertheless included it in category (121a).
89. Thus, it is even impossible to use a separate positive polar marker in the reduced adjective construction: in dat is wel waarschijnlijk or das ist schon wahrscheinlich 'that is probable' the polar marker can only be taken to have the adjective in its scope - e.g., as a denial of a denial that the state of affairs is probable - but not to combine with the adjective in expressing a positive polar epistemic evaluation of the state of affairs.
90. Andersson and Kvam also include a few instances of type (140) under this label. They consider the conjunctive particle to be part of the complement clause, yet separated from it by the complementing predicate. However, there is no formal, but only a semantic reason to consider the conjunction to belong to the complement clause rather than the clause with the mental state predicate, so strictly speaking these are somewhat different phenomena.
91. They also observe that the phenomenon, though most frequent with epistemic mental state predicates, is not limited to them, but also extends to some other types of complementing predicates, including communication, perception, and emotional attitude predicates.
92. The fact that the large majority of these semiparenthetical (and parenthetical) cases occurs in performative spoken contexts is symptomatic for the 'progressive' nature of this type of context as compared to others, in terms of flexibility in the use of grammar as well as of it being the prime source for language change (assuming that the development is from the full complementing via the semiparenthetical to the fully parenthetical pattern, of course).
93. Some Germans accept yet another form of a wh-question with a qualificational mental state predicate, viz. (cf. (150)) was glaubst Du warum ich rauche? 'what do you believe why I smoke?' (I have no corpus cases though; this structure is completely out of the question in English or Dutch). Unlike in the (formally) complementing cases, in this pattern the semantic position of the wh-word is clear, as it would be in the purely parenthetical case. But it avoids parentheticality, albeit at the cost of a quite strange syntactic pattern. Apparently, some speakers are willing to accept the latter in order to achieve the former.
94. 'Strongness' is obviously a relative notion, which may depend on factors such as personality, but also social status. Thus, what is felt to be strong will be determined to a large extent by the relative social position of speaker and listener: socially lower-ranking people will no doubt use this type of strategy more often relative to higher-ranking people than vice-versa.
95. An earlier, also corpus-based yet in several respects divergent analysis of the English modals was offered in Palmer (1979). On most points of divergence, Coates' analysis appears preferable, but I will not go into the details of this here.
96. Dendale (1994) argues that the French equivalent of must, devoir, is purely evidential, the epistemic meaning being an implication. But Tasmowski and Dendale (1994) argue for the same status for French pouvoir 'may'. For the Dutch and German equivalents, kunnen and können (see below), this is clearly not maintainable, however. As the analyses in this chapter will show, there are instances in my corpus data in which these modals are used to formulate conclusions from arguments or evidence presented in the preceding context, but this is far from a general rule. Intuitively, the same applies to English may.
97. The dynamic meaning involves inherent necessity, subject-internal (a need of the subject, comparable to inherent capacity), as in ik moet naar toilet 'I must go to the toilet', or situation-internal, as in je moet door de keuken om in de tuin te komen 'you must pass through the kitchen to get into the garden' (implying that there is no other way to do so). English must also appears to allow this meaning, even if Coates does not mention it (maybe because it is often very close to the deontic meaning, hence she may have subsumed instances under the latter meaning; Palmer 1979: 91ff does mention it as a distinct meaning category, though). The classical definition of dynamic modality usually only refers to 'ability' and 'inherent possibility', but this is clearly too narrow.
98. Sanders and Spooren (1996) also include predicates such as schijnen 'seem', lijken 'seem' and blijken 'appear' among the epistemic modals. However, although these are indeed modal-like predicates, they are purely evidential.
99. Sollen also has a hearsay-meaning which is lacking in should. But this meaning does belong under the label of evidentiality, of course.
100. Not all meaning change involves gain of a new meaning, of course, witness English do, which has developed from a full verb to a meaningless purely grammatical morpheme.
101. Bybee and Pagliuca (1985:67ff) link grammaticalization with semantic generalization in yet another way: the increase in grammaticalization in the development from dynamic and deontic to epistemic forms (see (c) below) is accompanied by a widening of scope from the agent to the entire proposition. This obviously involves a quite different sense of generality than bleaching as discussed above. And it is surely not related to degrees (let alone to the process) of grammaticalization per se, since the same difference exists between purely lexical expressions of these semantic categories. This goes to show the danger involved in focusing on just one expression type, a danger to which much of the grammaticalization literature is susceptible.
102. There is one modal adverb/adjective - possible/possibly (and its Dutch and German counterparts) - which can also express two qualificational categories, viz. dynamic and epistemic modality. But this is obviously not a systematic property of the modal adverbs and adjectives.
103. This is presupposing the view (Bybee et al. 1994, e.g.) that the inferential meaning of must developed out of the deontic one. This is not necessarily the case though: the inferential meaning could as well have developed straight out of the dynamic meaning, the deontic meaning being a parallel development (see Section 4.7; and see Goossens 1999 for evidence to this effect.) Even if so, this does not invalidate the principles of the above (and following) argumentation, however.
104. One may be inclined to make a similar claim for the development in lexical domains: to say that the development of the qualificational out of the non-qualificational meaning of the mental state predicates is a metaphorical mapping would seem to explain little (it essentially only restates the observation that there is a meaning change). The real question is: how did this change happen? It would be a surprise if this did not involve inferential processes as well.
105. Hence there were relatively more interpretational differences between myself and the second rater (see Section 1.6) for the Dutch modals than for the other expression types, and a considerable amount of joint deliberation was needed to come to the final classification.
106. The frequencies for Dutch kunnen in these tables deviate somewhat from those given in Nuyts (1996b), mainly due to the much larger amount of data taken into account here.
107. The relative frequencies of the different meanings of können in German found here are actually very different from those reported in Heine (1995). In particular, the present data show a considerably less prominent position for the epistemic meaning.
108. German könnte is the 'Konjunktiv II' form. There are also epistemic occurrences with 'Konjunktiv I', könne, viz. in the context of reported speech: Hans meint, diese Behauptung könne stimmen 'according to John this claim may be correct' (see Section 4.3 on such uses). This Konjunktiv I is then a marker of quotative evidentiality.
109. This may (partly) be why in you must be careful the deontic meaning predominates, while in you must be careless the evidential meaning first comes to mind (Wunderlich 1981): carefulness is an attitude one tends to recommend for the future, but carelessness is not. These are only strong tendencies, however, and one can easily construe contexts which overrule them.
110. If one confronts Dutch and German native speakers with cases such as in (201) or (199), or other similar ones to follow, as I have done, one consistently obtains inconsistent results. It is, of course, difficult to solicit reliable judgments of this kind anyway, given the subtlety of the matter (judging acceptability in combination with subtle meaning differences).
111. This expectation would actually not be borne out for the modals in their dynamic reading: intuitively they can be used focally; and see Section 5.4d. The question is whether the presumed difference in degree of grammaticalization between the dynamic and epistemic modals (cf. Section 4.1b) could account for any such information-structural difference. Maybe, the semantics of epistemic vs. dynamic modality as such is more important (see Section 6.5).
112. The relationship between negation and the modals in their different meanings, and specifically the question of how to understand the scope relationships between them, is a notoriously difficult problem. Palmer (1995, 1997), e.g., offers an analysis, with which I do not entirely agree (see also Nuyts 2000c), but discussion of this matter is for another occasion.
113. In the parallel pattern in Dutch (of which I have no corpus cases, though), one again uses the complementizer van instead of dat, i.e. het kan zijn van niet 'it may be of [i.e. that] not', as in the comparable pattern with a mental state predicate (cf. (117) in Section 3.4a).
114. This does not mean these other modals cannot have other types of strategic uses: Carretero (1997), e.g., argues that they do feature in politeness strategies.
115. It has been observed before, though never systematically investigated empirically (cf. Langacker 1974; Lötscher 1985; Chafe 1994) that, in terms of information structure, qualificational expressions in general (or 'regulatory units', as Chafe calls them) seem to behave differently from linguistic material representing the state of affairs (i.e. the 'objective content', as Langacker called it). When put in such general terms, however, this assumption is probably not entirely correct, or at least too unnuanced. See Section 6.5.
116. This is a tentative observation, though, since the MW and WNT are not optimal sources in this regards: as dictionaries, they offer reliable semantic information, but they may be less reliable for syntactic information.
117. Some arrows in (274) are tentative: e.g., from complementing mental state predicates to adverbs (or does this always go via parentheticals?), and from complementing mental state predicates to sentence-internal auxiliaries (or should this start from the parentheticals?). The schema obviously also disregards the fact that the epistemic auxiliaries (but not the mental state predicates) developed from independent predicates indirectly, via other modal meanings.
118. It is remarkable that this adverb, which derives from a phrase centrally involving mogen 'may', seems to have had an epistemic meaning right from the start, i.e. at least 4 centuries before the source modal itself. The other adverb and adjective deriving from mogen, mogelijk 'possibl(e/y)', however, also existed in Middle Dutch already, but with a predominant dynamic and secondary deontic meaning (which still exists, but is minor now), while the epistemic meaning seems to have developed around the same time as the epistemic meaning of the modal auxiliary.
119. The following is a summary overview. For more details, see the references given.
120. But probably there is a limit to the amount of information the situational network can contain, and if thematic units are too long, they may be handled in consecutive situational networks, whereby each is connected to the previous ones by the bookkeeping system (see below).
121. The present concept of a periphery in RRG has been adopted from earlier versions of the model, when there was no treatment of adverbials in terms of their scope yet, hence the entire set of optional constituents was simply considered to affect the core. In the logic of the new version it would probably be better to make the notion 'periphery' more global, covering all adverbials (as well as prepositional phrases) irrespective of the node to which they are attached. To signal their peripheral status, the attachment could be marked by a special type of link. Van Valin (pers.com.) appears inclined to accept such an approach.
122. The fact that FG, for good linguistic reasons, cannot integrate qualificational uses of predicative expression types in the layered system in a way which makes them comparable to semantically corresponding adverbials and operators (see Section 6.3b) further underscores the fact that linguistic structure is the wrong 'vehicle' to render the basic semantic dimensions involved.
123. Fillmore (1977) does not explicitly draw this consequence himself - his purpose was not the nature of conceptualization but the question how to develop a lexical semantics. But the conclusion is immanent in his entire theory of frame or scene semantics. Hence, he construes a mental model-like representation to render the conceptual scene of the commercial event. And see Fillmore (1985) for more explicit statements on the extra-linguistic status of conceptual frames.
124. Dik himself did not make clear statements regarding the conceptual status of qualifications. One element which might possibly contradict my assumption is that Dik assumes that the layered representation is constructed in the course of language generation in FG (cf. Section 6.3b), and some parts of it are built up even after syntactic function assignment (cf. Dik 1989a: 53, 1997a: 60), while syntactic functions are (obviously) explicitly not relevant for conceptual representation in Dik's view. Yet, as argued, the assumption that qualificational notions do not figure in human thought, but only get introduced when one starts speaking, would seem highly implausible.
125. Our observations clearly are a problem for the RRG decompositional semantic representation, since adverbials and operators are distinct elements in it. But I am not sure whether Van Valin would consider these representations to be identical to conceptual representations.
126. In this respect his proposal does not appear to reach a sufficient degree of generalization. But in other respects it does not seem sufficiently discriminating. Thus, Jackendoff's conceptual structures do not allow one to differentiate sell and give to X for Y, as in John sold the book to Mary for $50 vs. John gave the book to Mary for $50 (cf. Jackendoff 1990: 191, 194). Yet intuitively these are not fully equivalent in meaning: the latter does, but the former does not suggest that it was a bargain for Mary. One would expect to see this difference rendered in the conceptual representation underlying these forms, since, unlike the difference between buy and sell, it is not merely a matter of a different discourse perspective.
127. At least, the argument for an equal status is put forward in Jackendoff (1987: 193ff). Jackendoff (1983) still raised the impression that he considered his conceptual structures to be the representational format of long-term information, and even in Jackendoff (1987) and thereafter one sometimes finds statements which appear to suggest their predominance: e.g., “if long-term memories are stored predominantly in the form of conceptual structures …” (1987: 290). In any case, there can be no doubt that 3D and other representational formats (cf. his treatment of musical structure) take a very central role in Jackendoff's view of and concerns with human cognition. And there is also no doubt that Jackendoff is one of the few theorists so far who explicitly dares to venture into the way the multiple modalities of human perception and behavior diverge and relate, and should be integrated in a model of human cognition.
128. It is not hard to see that yet other representational systems are needed, next to an image-based structure, to grasp the ultimate meaning of yet other linguistic expressions: e.g., sour and sweet require recourse to olfactory representations, sticky to a tactile one, etc. See below.
129. Much or most knowledge of objects actually requires representations in the range of more than just one perceptual system. One not only knows what a typical cat looks like, but also how it smells, and how it feels. Similarly, one not only knows how garlic tastes, but also how it smells (and these two are in my experience far from identical, pace Jackendoff's 1987: 51f claim to the contrary), what it looks like, and how it feels if one touches it.
130. Such cases throw a shadow on an abstract kind of model of conceptual representation such as Johnson-Laird's (1983) mental models theory: it is not obvious at all how such a type of representation would be able to handle (generalizations over) non-vision-based knowledge. For, though mental models theory aims to be an abstract representational system (see Section 1.2a), it is still quite strongly related to imagistic representation.
131. One empirical way to investigate this would be to systematically go through a dictionary of a language and check whether there are any entries which refer to information which cannot be captured in terms of knowledge representation in the format of one of the basic perceptual and behavioral systems or generalizations thereof, and which do not (like epistemic modality and other qualificational categories) require recourse to an abstract level of representation to render their basic meaning in such a way that all linguistic instantiations of it can be grasped.
132. The different treatment of layering in RRG and FG is no doubt an effect of the different concept of levels in these models. RRG clearly differentiates between a fairly abstract semantic representation and a syntactic representation which is close to the surface of linguistic expression. FG, however, has only one central level of representation, the predicational, around which everything evolves, hence it is no surprise that this level should blend semantic and syntactic issues.
133. Since RRG has no full taxonomy of adverbials yet, it is not clear whether it would posit identical categories in the operator and adverbial ranges, or would rather allow for some differences between them, as in FG. As we will see, this issue has crucial consequences for the level at which the proposal can be interpreted, however.
134. One can, of course, also read this utterance with the time marker affecting the state of affairs only, and not the epistemic predicate. This would then be an example of a semiparenthetical use of the predicate, and in this case it obviously is performative.
135. Note that this means that temporal adverbs do not share the property of the modal adverbs that they cannot be focalized. See Section 6.5.3 below.
136. Note that this adds a further restriction to the view that modal adverbs only have scope over the newly asserted (focal) part of the utterance (see Section 1.4c): apparently, this does not apply, at least not straightforwardly, to other qualificational material in the sentence. Moreover, this observation at least complicates the assumption that (in languages such as Dutch or German) the clause-internal positioning of modal adverbs is determined by this scope, i.e. by what is newly asserted in the utterance (see Section 2.4h).
137. In the light of the foregoing, the RRG view (even the relativized one - see above) that in the operator range tense is higher than status might seem puzzling, particularly also in view of Bybee's (1985: 184) argument, specifically addressing RRG, that across languages of the world tense markers quite systematically occur closer to the predicate than grammatical markers of epistemic modality. The RRG proposal does correctly capture the mitigating use of past tense, of course, but clearly not the default time marking use of tense. The cause for this problem is no doubt that RRG in a language such as English automatically equates the position of tense with that of the finite verb. In an example such as he may be leaving soon it assumes that “tense is the left-most element; that is, because it is a suffix morphologically, it does not appear phonetically as the left-most element, but the syntax treats it as if it were left-most” (Van Valin and LaPolla 1997: 45). The critical question is, of course, whether may can be considered a tensed verb form at all. I have assumed it is not. And doing so brings English in line with Bybee's observations on languages which mark both time and epistemic modality by means of affixes.
138. On other occasions I have called this the 'hijacker principle' (e.g. Nuyts 1998).
139. For this kind of phenomenon, Construction Grammar as developed by Fillmore and colleagues (Fillmore 1988; Fillmore et al. 1988) allows a very elegant treatment, which might offer inspiration for modeling the lexicon in the kind of framework proposed here.
140. Note that I am talking about an 'emergent domain of grammar', applying Hopper's concept of 'emergence' to that specific domain only. I do not wish to imply that I subscribe to Hopper's general concept of Emergent Grammar. I do agree that “structure, or regularity, comes out of discourse and is shaped by discourse in an ongoing process” (Hopper 1998: 156). But, surely, there is more than only discourse to shape language, witness the discussion in Section 6.1. In fact, discourse itself is the product of an amalgam of dimensions, and is arguably emergent. For that reason alone, I disagree with the claim that grammar as a whole is “simply the name for certain categories of observed repetitions in discourse”, or “is, in other words, epiphenomenal”. Going into this is beyond the present concerns though.
141. This functional argument for decontextualized conceptual representation can be complemented with an argument on the basis of parsimony, and specifically of efficiency in knowledge storage. One chunk of knowledge may pertain to many different discourse situations, hence it would be quite inefficient if it were stored in a way specific for each context in which one has encountered it. For this would mean a potentially endless duplication in memory of essentially the same information. And it would moreover require a heavy conceptual processing load to keep track of how the different instantiations of the chunk of information relate. Moreover, as argued in Section 6.4 in connection with Dik's 'functional logic', to do the latter one probably needs a decontextualized concept of the chunk of knowledge anyway.
142. To the extent that language is just one of the means humans have to acquire and transmit knowledge (next to other communicative devices such as gesture, or direct perception, e.g. through vision), even the choice for language as a means is a 'contextualization' procedure.
143. Chafe (1994: 78) expresses reluctance regarding the usefulness of a notion of 'focus' (and 'topic') beyond notions such as contrastiveness and activation cost. Givón (1990: 699ff), on the other hand, does accept its relevance as a 'superficial cover term' for different deeper dimensions. I will not go into this discussion here, however, and continue to assume the latter position.
144. Language psychology obviously does admit certain types of flexibility, witness, e.g., the concept of 'incremental' processing which is now widely accepted in research on language production (cf. Kempen and Hoenkamp 1987; Levelt 1989). Yet, paradoxically, even granting its plausibility in principle, there appear limits to how far this particular kind of flexibility can be taken. Thus, how the above notions of integration vs. non-integration of expression forms and primary vs. secondary lexical selection of predicates can be translated in terms of such a concept is not obvious. But in any case, despite the fact that there are probably more (structurally) free elements in a clause than traditional linguistics would grant, even in spoken language most components of a clause - including the first produced ones - have a certain degree of grammatical cohesion. That is, they show different syntactic interdependency relations with some or all other components. An incremental model somehow has to deal with the difference between a predicate which is the head of its clause and one which is not, and between a constituent which is grammatically integrated in the clause and one which is not (irrespective of when in the production process it is actually uttered). It is not obvious how this can be done without assuming a full utterance plan (no matter how rough it may be). And since early as well as late constituents must fit into it, it would seem needed at an early stage in the production process.
145. RRG does assume a grouping of qualifications in syntax, of course, in that each qualification is related to one of the three major syntactic levels of structure postulated in RRG, viz. the nucleus, the core, and the clause. Hence each of those levels has several qualifications affecting it. However, it is far from obvious that such a linking is needed in syntactic structure at all: the linking is meant to handle the extension of scope of a qualification relative to (parts of) the information about the state of affairs, but that issue is no doubt conceptual again. See below.
146. I withdraw from my claim in Nuyts (1992a: 92) that the relative scope of time and deontic modality may vary. This was based on a mistaken interpretation of data.
147. Especially in view of the highly tentative nature of intuitions in these domains, checking the representativity of the following requires an in-depth empirical investigation both of the semantic categories of evidentiality and deontic modality (parallel to the present investigation of epistemic modality) and of the co-occurrences of expressions of these, mutually and with epistemic expressions. Such investigations would be highly desirable anyway: evidentiality (at least in West Germanic) and deontic modality (quite in general) have hardly received systematic attention (much less than epistemic modality), hence our current understanding of them is very poor. And a systematic empirical investigation of the co-occurrence possibilities and patterns of (types of expressions of) different qualificational categories is to my knowledge completely outstanding.
148. The question is whether the kind of combination in (283b) is really possible or would ever be used at all. Since the epistemic modals cannot be used descriptively, one might expect the same for the deontic ones. Determining whether this is true is something for a corpus investigation. However, one interesting observation is that Dutch, at least, has developed a syntactic pattern for the deontic modals which allows one to specify the source of the deontic evaluation, hence allows them to be used descriptively: e.g. Jan mag niet in de tuin van zijn mama (literally) 'John may not go into the garden from his mother', i.e. 'John's mother forbids him to go into the garden'. German has even developed a modal which, when used deontically, automatically implies that the deontic stance is due to someone else: Du sollst nach Hause gehen '(they request that) you must go home'. English does not seem to have similar strategies for the deontic modals. It does have forms such as John's mom says that he can't go into the garden, but this involves indirect speech reporting, which, as argued, is a very special context in terms of performativity and descriptivity.
149. The first person version of (285d), I seem to think that Mike will come to the party after all, may be conceivable in a highly peculiar context, viz. when the speaker realizes (s)he holds some view without having been aware of it. Even in that case, then, the mental state predicate is probably used descriptively, and not performatively: by means of it, the speaker does not express a commitment to an epistemic evaluation of the state of affairs, but, from a (mental) distance, describes an observation about his/her state of mind.
150. Taken in isolation, the strong mutual exclusivity of epistemic and evidential expressions might be considered an argument against treating epistemic modality and source evidentiality, or even evidentiality in general, as two separate qualificational categories, despite the arguments for doing so given in Section 1.3. For, after all, it is a quite general fact that different expression forms for one qualificational category cannot be combined in one utterance, a few exceptions involving harmonious forms aside. Yet the fact that the same mutual exclusivity exists for source-evidential and deontic, and for deontic and epistemic expressions unmasks this as a false argument.
151. If they were a case of borrowing, they would also go against the trend observed in Section 6.5 for forms to wander up the layering ladder.
152. I am assuming that adjectives such as plausible or logical can be classified as expressing types of evidentiality, viz. as involving gradations of inferentiality. See also Chapter 5.
153. One might object that this fact is obviously due to the absence of predicative time markers. For, as observed in the course of our analyses (and see Section 6.1), direct descriptive uses require a syntax which allows change in time or person, and that is only possible on predicative (verbal or adjectival) forms. But one can also turn this argument around and assume that the reason why time and lower qualifications do not, but higher qualifications do have predicative expressions is the result of a need for those higher qualifications, but not for the lower ones, to have a means to explicitly signal (absence of) speaker commitment to the qualification.
154. At first sight, one might be tempted to consider epistemic modality more complex than has been presented so far, too. For there is also the linguistic category which in the literature has been called realis vs. irrealis, which is not very prominent (though certainly not absent) in West Germanic, but is very manifest in many other languages of the world (cf., e.g., Givón 1994, 1995: 111ff; Chafe 1995; Mithun 1995), and which, like expressions of degrees of likelihood, is clearly concerned with the epistemic status of the state of affairs. However, this category can be taken to go back to the same epistemic scale as expressions of likelihood, but to express another dimension of it, viz. (roughly) whether the state of affairs as expressed in the utterance is conceptually marked for epistemic certainty or not. Obviously, investigating this linguistic category may throw interesting additional light on the conceptual structure of epistemic modality, but it need not lead to a split in the conceptual category as such.
155. Also at the linguistic level, however, this link between levels and function categories is disputable. First of all, the cluster 'representational + interpersonal' will not do as a typology of functions of language (see Nuyts 1992b: 26-64, 1993b). And secondly, as argued in several places above, there is no simple match between function categories and structural dimensions. The linguistic expression of any qualificational category involves an element of representation (of the conceptual qualification) and an element of adaptation to the interlocutors' concerns (the way the conceptual qualification is represented in function of the mutual positions - the intentions, relative social roles, etc. - of the speaker and the hearer in the communicative context).
156. Surely, both have a social dimension to them as well. What is good or bad is to a considerable extent socially determined (and acquired by an individual through socialization). And what can count as a valid inference is so as well. But in that way, everything conceptual is social or interpersonal. What one considers (too) fast or slow, or (too) large or small, is considerably determined by societal norms as well, yet aspectual categories - assuming qualifications of manner are such - are clearly very low in the layering system. Mind is a thoroughly social phenomenon, but that observation helps little to structure the layered system.
157. As mentioned in Section 1.3a, epistemic, deontic and dynamic modality have sometimes been characterized as relating a state of affairs to three different types of domains of knowledge. To the extent that this is meant to involve domains of external knowledge, this is no doubt fully adequate for deontic and epistemic modality (and probably for a few other qualificational dimensions as well), but much less so for dynamic modality: deontic and epistemic modality involve relating a state of affairs to knowledge which is entirely external to the information about the state of affairs proper, while dynamic modality involves a characterization of an observable feature internal to the state of affairs itself.
158. To what extent this means that an X̄-style syntax of the type introduced by Jackendoff (1977) and also adopted in some form in RRG is completely superfluous is a discussion for another occasion.
159. As a matter of fact, most or all qualificational dimensions which we have encountered at the sentence level (in main or subordinate domains) are also expressible within the noun phrase (cf. the level of the term operators in FG - see Section 6.3b). And qualificational expressions at that level appear at least roughly subject to the same organizational principles as those at the sentence level, microstructurally (i.e., their ordering being determined by semantic scope - cf. Rijkhoff 1990) and macrostructurally (in terms of the general rationale behind the organization; cf. Clark and Clark 1977: 474). Though going into this is beyond the present scope, one may expect that exactly the same layered system can be seen at work at the two levels of organization.
160. Apart from the elements to be discussed below, there are also plenty of mundane linguistic phenomena - also beyond epistemic modality and other qualificational categories - which appear to presuppose the existence of this control unit, as I have argued in Nuyts (1992b).
161. In a way, Jackendoff's view is understandable. Somehow it is true that one can get a good mental picture of how an utterance sounds, while this is not so for syntactic or conceptual structure. But this then probably involves an acoustic phonetic representation, and not phonological structure, as Jackendoff assumes. And the reason is probably that this phonetic structure still carries the vividness of one's direct percepts of the physical world, while levels further removed from the perceptive input lack this feature. Moreover, what would seem involved in this observation is 'imagery of perception': intuitively, when producing an utterance one does not get a better awareness of phonological (or even articulatory phonetic) structure than of syntactic or semantic structure at all.
162. Ahlqvist, A. (ed., 1982). Papers from the 5th international conference on historical linguistics. Amsterdam: Benjamins.
163. Aijmer, Karin (1997). I think: An English modal particle. In: T. Swan, O. J. Westvik (eds.), 1-47.
164. Akatsuka, Noriko (1985). Conditionals and the epistemic scale. Language 61: 625-639.
165. Aksu-Koç, Ayhan and Dan Slobin (1986). A psychological account of the development and use of evidentials in Turkish. In: W. Chafe, J. Nichols (eds.), 159-167.
166. Anderson, Lloyd B. (1986). Evidentials, paths of change, and mental maps: Typologically regular asymmetries. In: W. Chafe, J. Nichols (eds.), 273-312.
167. Andersson, Sven-Gunnar and Sigmund Kvam (1984). Satzverschränkung im heutigen Deutsch. Tübingen: Narr.
168. Auwera, Johan van der (1983). Misschien. Leuvense Bijdragen 72: 385-414.
169. Auwera, Johan van der and Vladimir A. Plungian (1998). Modality's semantic map. Linguistic Typology 2: 79-124.
170. Baars, Bernard J. (1988). A cognitive theory of consciousness. Cambridge: Cambridge University Press.
171. Baars, Bernard J. (1993). Putting the focus on the fringe: Three empirical cases. Consciousness and Cognition 2: 126-136.
172. Balota, David A. (1994). Visual word recognition: The journey from features to meaning. In: M. A. Gernsbacher (ed.), 303-358.
173. Barsalou, Lawrence W. and Christopher R. Hale (1993). Components of conceptual representation: From feature lists to recursive frames. In: I. Van Mechelen, J. Hampton, R. S. Michalski, P. Theuns (eds.), Categories and concepts. London: Academic Press, 97-144
174. Bartsch, Renate (1972). Adverbialsemantik. Frankfurt: Athenäum.
175. Belle, William Van (1973). Modale en performatieve werkwoorden. Spektator 3: 264-288.
176. Bellert, Irene (1977). On semantic and distributional properties of sentential adverbs. Linguistic Inquiry 8: 337-351.
177. Block, Ned (ed., 1981). Imagery. Cambridge, MA: MIT-Press.
178. Bock, Kathryn and Willem Levelt (1994). Language production: Grammatical encoding. In: M. A. Gernsbacher (ed.), 945-984.
179. Bock, Kathryn, Helga Loebell and Randal Morey (1992). From conceptual roles to structural relations: Bridging the syntactic cleft. Psychological Review 99: 150-171.
180. Bolinger, Dwight (1977). Meaning and form. London: Longman.
181. Bolinger, Dwight (1980). Wanna and the gradience of auxiliaries. In: G. Brettschneider, C. Lehmann (eds.), Wege zur Universalienforschung. Tübingen: Narr, 292-299.
182. Boogaart, P. C. Uit den (ed., 1975). Woordfrequenties in geschreven en gesproken Nederlands. Utrecht: Oosthoek.
183. Bouma, Lowell (1975). On contrasting the semantics of the modal auxiliaries of German and English. Lingua 37: 313-339.
184. Boyd, Julian and James Peter Thorne (1969). The semantics of modal verbs. Journal of Linguistics 5: 57-74.
185. Brandt, Margareta (1994). Subordination und Parenthese als Mittel der Informationsstrukturierung in Texten. Sprache und Pragmatik 1994: 1-37.
186. Bruner, Jerome (1990). Acts of meaning. Cambridge, MA: Harvard University Press.
187. Brünner, Gisela (1981). Modalverben und die Realisierung von Sprechhandlungen. In: I. Rosengren (ed.), 91-101.
188. Bühler, Karl (1934). Sprachtheorie. Stuttgart: Fischer.
189. Burge, Tyler (1977). Belief de re. Journal of Philosophy 74: 338-362.
190. Bustos, E. de (1995). On the very idea of cognitivism. Pragmatics & Cognition 3: 147-157.
191. Bybee, Joan L. (1985). Morphology. Amsterdam: Benjamins.
192. Bybee, Joan L. (1988). Semantic substance vs. contrast in the development of grammatical meaning. Berkeley Linguistic Society 14: 247-264.
193. Bybee, Joan L. (1995). The semantic development of past tense modals in English. In: J. Bybee, S. Fleischman (eds.), 503-517.
194. Bybee, Joan L. and Suzanne Fleischman (eds., 1995). Modality in grammar and discourse. Amsterdam: Benjamins.
195. Bybee, Joan L. and William Pagliuca (1985). Cross-linguistic comparison and the development of grammatical meaning. In: J. Fisiak (ed.), Historical semantics - historical word-formation. Berlin: Mouton, 59-83.
196. Bybee, Joan L. and Wiliam Pagliuca (1987). The evolution of future meaning. In: A. G. Ramat, O. Carruba, G. Bernini (eds.), Papers from the 7th international conference on historical linguistics. Amsterdam: Benjamins, 109-122.
197. Bybee, Joan L., William Pagliuca and Revere D. Perkins (1991). Back to the future. In: E. C. Traugott, B. Heine (eds.), 17-58.
198. Bybee, Joan L., Revere D. Perkins and William Pagliuca (1994). The evolution of grammar: Tense, aspect and modality in the languages of the world. Chicago: University of Chicago Press.
199. Calbert, Joseph P. (1975). Toward the semantics of modality. Studien zur deutschen Grammatik 1: 1-70.
200. Carey, Kathleen (1995). Subjectification and the development of the English perfect. In: D. Stein, S. Wright (eds.), 83-102.
201. Carretero, Marta (1996). A new angle on an old theme: Epistemic modality reconsidered. Estudios Ingleses de la Universidad Complutense 4: 253-263.
202. Carretero Lapeyre, Marta (1997). The relevance of politeness in the epistemic interpretation of the English modals. Pragmalingüística 3/4: 241-259.
203. Chafe, Wallace (1986). Evidentiality in English conversation and academic writing. In: W. Chafe, J. Nichols (eds.), 261-272.
204. Chafe, Wallace (1994). Discourse, consciousness, and time. Chicago: University of Chicago Press.
205. Chafe, Wallace (1995). The realis-irrealis distinction in Caddo, the Northern Iroquoian languages, and English. In: J. Bybee, S. Fleischman (eds.), 349-365.
206. Chafe, Wallace (1996). How consciousness shapes language. Pragmatics & Cognition 4: 35-54.
207. Chafe, Wallace and Johanna Nichols (eds., 1986). Evidentiality. Norwood: Ablex.
208. Choi, Soonja (1995). The development of epistemic sentence-ending modal forms and functions in Korean children. In: J. Bybee, S. Fleischman (eds.), 165-204.
209. Clark, Herbert H. (1992). Arenas of language use. Chicago: University of Chicago Press.
210. Clark, Herbert H. (1996). Using language. Cambridge: Cambridge University Press.
211. Clark, Herbert H. and Eve V. Clark (1977). Psychology and language. New York: Harcourt.
212. Coates, Jennifer (1983). The semantics of the modal auxiliaries. London: Croom Helm.
213. Coates, Jennifer (1987). Epistemic modality and spoken discourse. Transactions of the Philological Society (Oxford) 1987: 110-131.
214. Daems, Frans and Louis Goossens (eds., 1983), Een spyeghel voor G. Jo Steenbergen. Leuven: Acco.
215. Dascal, Marcelo (1996). The dispute on the primacy of thinking or speaking. In: M. Dascal, D. Gerhardus, K. Lorenz, G. Meggle (eds.), Sprachphilosophie, vol. 2. Berlin: De Gruyter, 1024-1041.
216. DeLancey, Scott (1986). Evidentiality and volitionality in Tibetan. In: W. Chafe, J. Nichols (eds.), 203-213.
217. DeLancey, Scott (1997). Mirativity: The grammatical marking of unexpected information. Linguistic Typology 1: 33-52.
218. Dendale, Patrick (1994). Devoir épistémique, marqueur modal ou évidentiel? Langue Française 102: 24-40.
219. Dennett, Daniel C. (1991). Consciousness explained. New York: Little.
220. Dietrich, Rainer (1992). Modalität im Deutschen. Opladen: Westdeutscher Verlag.
221. Dijk, Teun A. van and Walter Kintsch (1983). Strategies of discourse comprehension. New York: Academic Press.
222. Dik, Simon C. (1978). Functional Grammar. Amsterdam: North-Holland.
223. Dik, Simon C. (1986). On the notion 'functional explanation'. Belgian Journal of Linguistics 1: 11-52.
224. Dik, Simon C. (1987a). Linguistically motivated knowledge representation. In: M. Nagao (ed.), Language and artificial intelligence. Amsterdam: North-Holland, 145-170.
225. Dik, Simon C. (1987b). Functional grammar and its potential computer applications. In: W. Meijs (ed.), Corpus linguistics and beyond. Amsterdam: Rodopi, 253-268.
226. Dik, Simon C. (1988). Concerning the logical component of a natural language generator. In: M. Zock, G. Sabah (eds.), Advances in natural language generation, vol. 1. London: Pinter, 73-91.
227. Dik, Simon C. (1989a). The theory of functional grammar, pt. 1. Dordrecht: Foris.
228. Dik, Simon C. (1989b). Towards a unified cognitive language. In: F. Heyvaert, F. Steurs (eds.), Worlds behind words. Leuven: Leuven University Press, 97-110.
229. Dik, Simon C. (1990). On the semantics of conditionals. In: J. Nuyts et al. (eds.), 233-261.
230. Dik, Simon C. (1997a). The theory of functional grammar, pt. 1: The structure of the clause. Berlin: De Gruyter.
231. Dik, Simon C. (1997b). The theory of functional grammar, pt. 2: Complex and derived constructions. Berlin: De Gruyter.
232. Dik, Simon C. and Kees Hengeveld (1991). The hierarchical structure of the clause and the typology of perception verb complements. Linguistics 29: 231-259.
233. Dik, Simon C., Kees Hengeveld, Elseline Vester and Co Vet (1990). The hierarchical structure of the clause and the typology of adverbial satellites. In: J. Nuyts et al. (eds.), 25-70.
234. Dittmar, Norbert and Astrid Reich (eds., 1993). Modality in language acquisition. Berlin: De Gruyter.
235. Doherty, Monika (1985). Epistemische Bedeutung. Berlin: Akademie Verlag.
236. Duinhoven, A. M. (1994). Over modaliteit gesproken. De Nieuwe Taalgids 87: 30-56.
237. Egan, Owen (1986). The concept of belief in cognitive theory. In: L. P. Mos (ed.), Annals of theoretical psychology, vol. 4. New York: Plenum, 315-350.
238. Ehrman, M. (1966). The meaning of the modals in present-day American English. The Hague: Mouton.
239. Ellis, Ralph D. (1995). Questioning consciousness. Amsterdam: Benjamins.
240. Falkenberg, G. (ed., 1989). Wissen, Wahrnehmen, Glauben. Tübingen: Niemeyer.
241. Fillmore, Charles J. (1968). The case for case. In: E. Bach, R. Harms (eds.), Universals in linguistic theory. London: Holt, 1-88.
242. Fillmore, Charles J. (1977). Topics in lexical semantics. In: R. W. Cole (ed.), Current issues in linguistic theory. Bloomington: Indiana University Press, 76-138.
243. Fillmore, Charles J. (1985). Frames and the semantics of understanding. Quaderni di Semantica 6: 222-254.
244. Fillmore, Charles J. (1988). The mechanisms of 'construction grammar'. Berkeley Linguistic Society 14: 35-55.
245. Fillmore, Charles, Paul Kay and M. C. O'Connor (1988). Regularity and idiomaticity in grammatical construction: The case of 'let alone'. Language 64: 501-538.
246. Fleischman, Suzanne (1989). Temporal distance: A basic linguistic metaphor. Studies in Language 13: 1-51.
247. Fletcher, Charles R. (1994). Levels of representation in memory for discourse. In: M. A. Gernsbacher (ed.), 589-607.
248. Fodor, Jerry A. (1968). Psychological explanation. New York: Random.
249. Fodor, Jerry A. (1975). The language of thought. New York: Crowell.
250. Fodor, Jerry A. (1983). The modularity of mind. Cambridge, MA: MIT-Press.
251. Fodor, Jerry A. (1987). Psychosemantics. Cambridge, MA: MIT-Press.
252. Foley, William A. and Robert D. Van Valin (1984). Functional syntax and universal grammar. Cambridge: Cambridge University Press.
253. Fortescue, Michael, Peter Harder and Lars Kristoffersen (eds., 1992). Layered structure and reference in a functional perspective. Amsterdam: Benjamins.
254. Frajzyngier, Zygmunt (1991). The de dicto domain in language. In: E. Traugott, B. Heine (eds.), vol. 1, 219-251.
255. Frajzyngier, Zygmunt (1995). A functional theory of complementizers. In: J. Bybee, S. Fleischman (eds.), 473-502.
256. Gamon, D. (1994). On the development of epistemicity in the German modal verbs mögen and müssen. Folia Linguistica Historica 14: 125-176.
257. Gernsbacher, Morton Ann (ed., 1994). Handbook of psycholinguistics. San Diego: Academic Press.
258. Givón, Talmy (1982). Evidentiality and epistemic space. Studies in Language 6: 23-49.
259. Givón, Talmy (1984). Syntax: A functional-typological introduction, vol. 1. Amsterdam: Benjamins.
260. Givón, Talmy (1990). Syntax: A functional-typological introduction, vol. 2. Amsterdam: Benjamins.
261. Givón, Talmy (1994). Irrealis and the subjunctive. Studies in Language 18: 265-337.
262. Givón, Talmy (1995). Functionalism and grammar. Amsterdam: Benjamins.
263. Givón, Talmy (1998). The functional approach to grammar. In: M. Tomasello (ed.), 41-66.
264. Glenberg, Arthur M., Peter Kruley and William E. Langston (1994). Analogical processes in comprehension: Simulation of a mental model. In: M. A. Gernsbacher (ed.), 609-640.
265. Goossens, Louis (1982). On the development of the modals and of the epistemic function in English. In: A. Ahlqvist (ed.), 74-84.
266. Goossens, Louis (1983). Can and kunnen: Dutch and English potential compared. In: F. Daems, L. Goossens (eds.), 147-158.
267. Goossens, Louis (1985a). Modality and the modals. In: M. Bolkestein, C. De Groot, L. Mackenzie (eds.), Predicates and terms in functional grammar. Dordrecht: Foris, 203-217.
268. Goossens, Louis (1985b). Differentiating the English modals in functional grammar. In: J. Nuyts (ed.), Antwerp studies in functional grammar. Wilrijk: Antwerp Papers in Linguistics 39, 49-71.
269. Goossens, Louis (1987). The auxiliarization of the English modals: A functional grammar view. In: M. Harris, P. Ramat (eds.), Historical development of auxiliaries. Berlin: Mouton, 111-143.
270. Goossens, Louis (1990). Cunnan, conne(n), can: The development of a radial category. In: G. Kellermann, M. Morrissey (eds.), Diachrony within synchrony: Language history and cognition. Frankfurt/Mn: Lang, 377-394.
271. Goossens, Louis (1999). Metonymic bridges in modal shifts. In: K.-U. Panther, G. Radden (eds.), Metonymy in language and cognition. Amsterdam: Benjamins, 193-210.
272. Greenbaum, Sidney (1969). Studies in English adverbial usage. London: Longman.
273. Greenbaum, Sidney, Geoffrey Leech and Jan Svartvik (eds., 1979). Studies in English linguistics for Randolph Quirk. London: Longman.
274. Grice, H. P. (1969). Utterer's meaning and intentions. Philosophical Review 78: 147-177.
275. Gumperz, John J. and Stephen C. Levinson (eds., 1996). Rethinking linguistic relativity. Cambridge: Cambridge University Press.
276. Guo, Jiansheng (1995). The interactional basis of the Mandarin modal néng 'can'. In: J. Bybee, S. Fleischman (eds.), 205-238.
277. Haas-Spohn, Ulrike (1989). Zur Interpretation der Einstellungszuschreibungen. In: G. Falkenberg (ed.), 49-94.
278. Haiman, John (ed., 1985). Iconicity in syntax. Amsterdam: Benjamins.
279. Halliday, Michael A. K. (1976). System and function in language. London: Oxford University Press.
280. Halliday, Michael A. K. (1994). An introduction to functional grammar. London: Arnold.
281. Hanson, Kristin (1987). On subjectivity in the history of epistemic expressions in English. Chicago Linguistic Society 23: 133-147.
282. Harder, Peter (1996). Functional semantics. Berlin: De Gruyter.
283. Hayes-Roth, Barbara (1985). A blackboard architecture for control. Artificial Intelligence 26: 251-321.
284. Heine, Bernd (1993). Auxiliaries: Cognitive forces and grammaticalization. Oxford: Oxford University Press.
285. Heine, Bernd (1995). Agent-oriented vs. epistemic modality: Some observations on German modals. In: J. Bybee, S. Fleischman (eds.), 17-53.
286. Heine, Bernd, Ulrike Claudi and Friederike Hünnemeyer (1991). Grammaticalization: A conceptual framework. Chicago: The University of Chicago Press.
287. Hengeveld, Kees (1988). Illocution, mood and modality in a functional grammar of Spanish. Journal of Semantics 6: 227-269.
288. Hengeveld, Kees (1989). Layers and operators in functional grammar. Journal of Linguistics 25: 127-157.
289. Hengeveld, Kees (1990). The hierarchical structure of utterances. In: J. Nuyts et al. (eds.), 1-23.
290. Hengeveld, Kees (1992a). Non-verbal predication. Berlin: De Gruyter.
291. Hengeveld, Kees (1992b). Parts of speech. In: M. Fortescue et al. (eds.), 29-55.
292. Hickmann, Maya, Christian Champaud and Dominique Bassano (1993). Pragmatics and metapragmatics in the development of epistemic modality: Evidence from French children's reports of think-statements. First Language 13: 359-389.
293. Hintikka, Jaakko (1975). The intentions of intentionality and other new models for modalities. Dordrecht: Reidel.
294. Hofmann, Thomas R. (1976). Past tense replacement and the modal system. In: J. McCawley (ed.), Syntax and semantics, vol. 7: Notes from the linguistic underground. New York: Academic Press, 85-100.
295. Hopper, Paul J. (1987). Emergent grammar. Berkeley Linguistic Society 13: 139-157.
296. Hopper, Paul J. (1991). On some principles of grammaticization. In: E. Traugott, B. Heine (eds.), vol. 1, 17-35.
297. Hopper, Paul J. (1998). Emergent grammar. In: M. Tomasello (ed.), 155-175.
298. Hopper, Paul J. and Elizabeth Closs Traugott (1993). Grammaticalization. Cambridge: Cambridge University Press.
299. Hoye, Leo (1997). Adverbs and modality in English. London: Longman.
300. Huddleston, Rodney (1976). Some theoretical issues in the description of the English verb. Lingua 40: 331-383.
301. Huddleston, Rodney (1979). Criteria for auxiliaries and modals. In: S. Greenbaum et al (eds.), 65-78.
302. Jackendoff, Ray (1972). Semantic interpretation in generative grammar. Cambridge, MA: MIT-Press.
303. ¯ syntax: A study of phrase structure. Cambridge, MA: MIT-Press.
304. Jackendoff, Ray (1983). Semantics and cognition. Cambridge, MA: MIT-Press.
305. Jackendoff, Ray (1987). Consciousness and the computational mind. Cambridge, MA: MIT-Press.
306. Jackendoff, Ray (1990). Semantic structures. Cambridge, MA: MIT-Press.
307. Jackendoff, Ray (1996). How language helps us think. Pragmatics & Cognition 4: 1-34.
308. Johnson, Mark (1987). A new approach to clause structure in role and reference grammar. Davis Working Papers in Linguistics 2: 55-59.
309. Johnson-Laird, Philip N. (1983). Mental models. Cambridge: Cambridge University Press.
310. Kamp, Hans and Uwe Reyle (1993). From discourse to logic. Dordrecht: Kluwer.
311. Kay, Paul (1996). Intra-speaker relativity. In: J. Gumperz, S. Levinson (eds.), 97-114.
312. Keller, Charles M. and Janet Dixon Keller (1996). Imaging in iron, or thought is not inner speech. In: J. Gumperz, S. Levinson (eds.), 115-129.
313. Kempen, Gerard and Edward Hoenkamp (1987). An incremental procedural grammar for sentence formulation. Cognitive Science 11: 201-258.
314. Kiefer, Ferenc (1984). Focus and modality. Groninger Arbeiten zur Germanistischen Linguistik 24: 55-81.
315. Kintsch, Walter (1994). The psychology of discourse processing. In: M. A. Gernsbacher (ed.), 721-739.
316. Kiparsky, P. and C. Kiparsky (1970). Fact. In: M. Bierwisch, K. Heidolph (eds.), Progress in linguistics. The Hague: Mouton, 143-173.
317. Kirsner, Robert and Sandra A. Thompson (1976). The role of pragmatic inference in semantics: A study of sensory verb complements in English. Glossa 10: 200-240.
318. Kita, Sotaro (1993). Language and thought interface: A study of spontaneous gestures and Japanese mimetics. University of Chicago Ph. D. dissertation.
319. Kosslyn, Stephen M. (1980). Image and mind. Cambridge, MA: Harvard University Press.
320. Kosslyn, Stephen M. (1994). Image and brain. Cambridge, MA: MIT-Press.
321. Kratzer, Angelika (1978). Semantik der Rede. Königstein: Scriptor.
322. Kress, G. R. (1977). Tense as modality. University of East Anglia Papers in Linguistics 3: 40-52.
323. Lakoff, George (1970). Linguistics and natural logic. Synthese 22: 151-271.
324. Lakoff, George (1987). Women, fire, and dangerous things. Chicago: University of Chicago Press.
325. Lakoff, Robin T. (1972). The pragmatics of modality. Chicago Linguistic Society 8: 229-246.
326. Lambrecht, Knud (1994). Information structure and sentence form. Cambridge: Cambridge University Press.
327. Lang, Ewald (1979). Zum Status der Satzadverbiale. Slovo a Slovenost 40: 200-213.
328. Lang, Ewald and Renate Steinitz (1977). Können Satzadverbiale performativ gebraucht werden? Linguistische Studien A-42: 51-96.
329. Langacker, Ronald (1974). Movement rules in functional perspective. Language 50: 630-664.
330. Langacker, Ronald (1987). Foundations of cognitive grammar, pt. 1. Stanford: Stanford University Press.
331. Langacker, Ronald (1990). Subjectification. Cognitive Linguistics 1: 5-38.
332. Langacker, Ronald (1997). The contextual basis of cognitive semantics. In: J. Nuyts, E. Pederson (eds.), 229-252.
333. Lee, Hyo Sang (1993). Cognitive constraints on expressing newly perceived information: With reference to epistemic modal suffixes in Korean. Cognitive Linguistics 4: 135-167.
334. Leech, Geoffrey and Jennifer Coates (1979). Semantic indeterminacy and the modals. In: S. Greenbaum et al. (eds.), 79-90.
335. Levelt, Willem J. M. (1989). Speaking. Cambridge, MA: MIT-Press.
336. Levinson, Stephen C. (1983). Pragmatics. Cambridge: Cambridge University Press.
337. Levinson, Stephen C. (1996a). Language and space. Annual Review of Anthropology 25: 353-382.
338. Levinson, Stephen C. (1996b). Relativity in spatial conception and description. In: J. Gumperz, S. Levinson (eds.), 177-202.
339. Levinson, Stephen C. (1997). From outer to inner space: Linguistic categories and non-linguistic thinking. In: J. Nuyts, E. Pederson (eds.), 13-45.
340. Linell, Per (1982). The written language bias in linguistics. Linköping: University of Linköping Press.
341. Lively, Scott, David Pisoni and Stephen Goldinger (1994). Spoken word recognition: Research and theory. In: M. A. Gernsbacher (ed.), 265-301.
342. Lötscher, Andreas (1985). Akzentuierung und Thematisierbarkeit von Angaben. Linguistische Berichte 97: 228-251.
343. Lucy, John A. (1992). Language diversity and thought. Cambridge: Cambridge University Press.
344. Lunn, Patricia V. (1995). The evaluative function of the Spanish subjunctive. In: J. Bybee, S. Fleischman (eds.), 429-449.
345. Lyons, John (1977). Semantics. Cambridge: Cambridge University Press.
346. Mangan, Bruce (1993). Taking phenomenology seriously: The 'fringe' and its implications for cognitive research. Consciousness and Cognition 2: 89-108.
347. Marr, David (1982). Vision. San Francisco: Freeman.
348. McGlone, Matthew S. and Ann B. Reed (1998). Anchoring in the interpretation of probability expressions. Journal of Pragmatics 30: 723-733.
349. McNeill, David (1992). Hand and mind: What gestures reveal about thought. Chicago: University of Chicago Press.
350. McNeill, David (1997). Growth points cross-linguistically. In: J. Nuyts, E. Pederson (eds.), 190-212.
351. Miller, George A. and Donna M. Kwilosz (1981). Interactions of modality and negation in English. In: A. Joshi, B. Webber, I. Sag (eds.), Elements of discourse understanding. Cambridge: Cambridge University Press, 201-216.
352. Mitchell, Don C. (1994). Sentence parsing. In: M. A. Gernsbacher (ed.), 375-409.
353. Mithun, Marianne (1995). On the reality of irreality. In: J. Bybee, S. Fleischman (eds.), 367-388.
354. Mudersbach, Klaus (1984). Kommunikation über Glaubensinhalte: Grundlagen der epistemischen Linguistik. Berlin: De Gruyter.
355. Natsoulas, Thomas (1983). Concepts of consciousness. The Journal of Mind and Behavior 4: 13-59.
356. Nichols, Johanna (1986). The bottom line: Chinese Pidgin Russian. In: W. Chafe, J. Nichols (eds.), 239-257.
357. Noël, Dirk (1997). The choice between infinitives and that-clauses after believe. English Language and Linguistics 1: 271-284.
358. Noonan, Michael (1985). Complementation. In: T. Shopen (ed.), Language typology and syntactic description, pt. 2: Complex constructions. Cambridge: Cambridge University Press, 42-140.
359. Nordlinger, Rachel and Elizabeth Closs Traugott (1997). Scope and the development of epistemic modality: Evidence from ought to. English Language and Linguistics 1: 295-317.
360. Nuyts, Jan (1989). Functional procedural grammar: An overview. Amsterdam: Working Papers in Functional Grammar 31.
361. Nuyts, Jan (1990a). Negative-raising reconsidered. Journal of Pragmatics 14: 559-588.
362. Nuyts, Jan (1990b). Emotions and the functionality of language. Grazer Linguistische Studien 33/34: 227-240.
363. Nuyts, Jan (1990c). Linguistic representation and conceptual knowledge representation. In: J. Nuyts et al. (eds.), 263-293.
364. Nuyts, Jan (1992a). Subjective vs. objective modality: What is the difference? In: M. Fortescue et al. (eds.), 73-98.
365. Nuyts, Jan (1992b). Aspects of a cognitive-pragmatic theory of language. Amsterdam: Benjamins.
366. Nuyts, Jan (1993a). Epistemic modal adverbs and adjectives and the layered representation of conceptual and linguistic structure. Linguistics 31: 933-969.
367. Nuyts, Jan (1993b). On determining the functions of language. Semiotica 94: 201-232.
368. Nuyts, Jan (1993c). From language to conceptualization: The case of epistemic modality. Chicago Linguistic Society 29, vol. 2 (parasession: What we think, what we mean, and how we say it): 271-286.
369. Nuyts, Jan (1993d). Cognitive linguistics. Journal of Pragmatics 20: 269-290.
370. Nuyts, Jan (1994). The intentional and the socio-cultural in language use. Pragmatics & Cognition 2: 237-268.
371. Nuyts, Jan (1995a). Afterthoughts to E. de Bustos (1995), 'On the very idea of cognitivism'. Pragmatics & Cognition 3: 279-282.
372. Nuyts, Jan (1995b). Functionalism vs. formalism. In: J. Verschueren, J.-O. Östman, J. Blommaert (eds.), Handbook of pragmatics: Manual. Amsterdam: Benjamins, 293-300.
373. Nuyts, Jan (1996a). Consciousness in language. Pragmatics & Cognition 4: 153-180.
374. Nuyts, Jan (1996b). Das modale Hilfsverb kunnen und das System der epistemischen Modalität im Niederländischen: Überlegungen zur Synchronie und Diachronie. Leuven Contributions to Linguistics and Philology 85: 33-54.
375. Nuyts, Jan (1996c). Discourse and conceptualization in expressing epistemic modality. In: R Meyer-Klabunde, C. von Stutterheim (eds.), Conceptual and semantic knowledge in language production, vol. 2. Heidelberg: Arbeiten aus dem SFB 245 Sprache und Situation, 235-251.
376. Nuyts, Jan (1997). How do you think? In: C. Butler, J. Connolly, R. Gatward, R. Vismans (eds.), A fund of ideas. Amsterdam: IFOTT, 3-18.
377. Nuyts, Jan (1998). Layered models of qualifications of states of affairs: Cognition vs. typology? In: J. van der Auwera, F. Durieux, L. Lejeune (eds.), English as a human language. München: Lincom Europa, 274-284.
378. Nuyts, Jan (1999). Processing epistemic modality: Some consequences for modeling language production. In: R. Meyer-Klabunde, C. von Stutterheim (eds.), Representations and processes in language production. Wiesbaden: Westdeutscher Verlag, 181-203.
379. Nuyts, Jan (2000a). Tensions between discourse structure and conceptual semantics: The syntax of epistemic modal expressions. Studies in Language 24: 103-135.
380. Nuyts, Jan (2000b). Subjectivity as an evidential dimension in epistemic modal expressions. Journal of Pragmatics, in press.
381. Nuyts, Jan (2000c). Negatie en het epistemische gebruik van de modale hulpwerkwoorden in het Nederlands. In: S. Gillis, J. Nuyts, J. Taeldeman (eds.), Met taal om de tuin geleid: Opstellen voor Georges De Schutter. Wilrijk: UIA, 257-267.
382. Nuyts, Jan and Eric Pederson (eds., 1997). Language and conceptualization. Cambridge: Cambridge University Press.
383. Nuyts, Jan and Wietske Vonk (1999). Epistemic modality and focus in Dutch. Linguistics 37, 699-737.
384. Nuyts, Jan, A. Machtelt Bolkestein and Co Vet (eds., 1990). Layers and levels of representation in language theory. Amsterdam: Benjamins.
385. Öhlschläger, Günther (1989). Zur Syntax und Semantik der Modalverben des Deutschen. Tübingen: Niemeyer.
386. Paivio, A. (1972). Imagery and verbal processes. New York: Holt.
387. Paivio, A. (1991). Images in mind. New York: Harvester.
388. Palmer, F. R. (1965). A linguistic study of the English verb. London: Longman.
389. Palmer, F. R. (1979). Modality and the English modals. London: Longman.
390. Palmer, F. R. (1983). Semantic explanations for the syntax of the English modals. In: F. Heny, B. Richards (eds.), Linguistic categories, vol. 2. Dordrecht: Reidel, 205-217.
391. Palmer, F. R. (1986). Mood and modality. Cambridge: Cambridge University Press.
392. Palmer, F. R. (1995). Negation and the modals of of possibility and necessity. In: J. Bybee, S. Fleischman (eds.), 453-471.
393. Palmer, Frank R. (1997). Negation and modality in the Germanic languages. In: T. Swan, O. J. Westvik (eds.), 133-149.
394. Pederson, Eric and Jan Nuyts (1997). On the relationship between language and conceptualization. In: J. Nuyts, E. Pederson (eds.), 1-12.
395. Perkins, Michael R. (1982). The core meanings of the English modals. Journal of Linguistics 18: 245-273.
396. Perkins, Michael R. (1983). Modal expressions in English. London: Pinter.
397. Pinker, Steven (1994). The language instinct. New York: Morrow.
398. Plank, Frans (1981). Modalitätsausdruck zwischen Autonomie und Auxiliarität. In: I. Rosengren (ed.), 57-71.
399. Plank, Frans (1984). The modals story retold. Studies in Language 8: 305-364.
400. Premper, Waldfried (1991). Propositionen, positionale Operatoren und Situierung. AKUP: Arbeiten des Kölner Universalien-Projekts 85: 1-41.
401. Pylyshyn, Zenon W. (1973). What the mind's eye tells the mind's brain: A critique of mental imagery. Psychological Bulletin 80: 1-24.
402. Pylyshyn, Zenon W. (1984). Computation and cognition. Cambridge, MA: MIT-Press.
403. Quine, W. V. O. (1956). Quantifiers and propositional attitudes. Journal of Philosophy 53: 177-187.
404. Ramat, Paolo and Davide Ricca (1994). Prototypical adverbs: On the scalarity/radiality of the notion 'adverb'. Rivista di Linguistica 6: 289-326.
405. Ramat, Paolo and David Ricca (1998). Sentence adverbs in the languages of Europe. In: J. van der Auwera (ed.), Adverbial constructions in the languages of Europe. Berlin: Mouton, 187-275.
406. Ransom, Evelyn (1977). On the representation of modality. Linguistics and Philosophy 1: 357-379.
407. Ransom, Evelyn (1986). Complementation: Its meanings and forms. Amsterdam: Benjamins.
408. Rijkhoff, Jan (1990). Toward a unified analysis of terms and predications. In: J. Nuyts et al. (eds.), 165-191.
409. Rolf, Eckard (1989). Zur (vermeintlichen) Ambiguität der Einstellungssätze. In: G. Falkenberg (ed.), 175-196.
410. Rosengren, Inger (ed., 1981). Sprache und Pragmatik: Lundner Symposium 1980. Lund: CWK Gleerup.
411. Ross, John R. (1970). On declarative sentences. In: R. Jacobs, P. Rosenbaum (eds.), Readings in English transformational grammar. Waltham: Ginn, 222-272.
412. Sanders, José and Wilbert Spooren (1996). Subjectivity and certainty in epistemic modality: A study of Dutch epistemic modifiers. Cognitive Linguistics 7: 241-264.
413. Sanford, Anthony J. and Simon C. Garrod (1994). Selective processing in text understanding. In: M. A. Gernsbacher (ed.), 699-719
414. Schank, Roger C., N. M. Goldman, C. J. Rieger III and C. K. Riesbeck (1975). Conceptual information processing. Amsterdam: North Holland.
415. Schneider, Walter and Richard M. Shiffrin (1977). Controlled and automatic human information processing, I: Detection, search and attention. Psychological Review 84: 1-66.
416. Schutter, Georges De (1983). Modaliteit en andere modificaties in de Nederlandse grammatica. In: F. Daems, L. Goossens (eds.), 277-291.
417. Schutter, Georges De and P. van Hauwermeiren (1983). De structuur van het Nederlands: Taalbeschouwelijke grammatica. Malle: De Sikkel.
418. Schutter, Georges De and Jan Nuyts (1983). Toward an integrated model of a functional grammar. In: S. Dik (ed.), Advances in functional grammar. Dordrecht: Foris, 387-404.
419. Searle, John (1972). Chomsky's revolution in linguistics. The New York Review of Books, June 29: 16-24.
420. Searle, John (1983). Intentionality. Cambridge: Cambridge University Press.
421. Seuren, Pieter A. M. (1969). Operators and nucleus. Cambridge: Cambridge University Press.
422. Shepherd, Susan C. (1982). From deontic to epistemic. In: A. Ahlqvist (ed.), 316-323.
423. Shepherd, Susan C. (1993). The acquisition of modality in Antiguan Creole. In: N. Dittmar, A. Reich (eds.), 171-184.
424. Shiffrin, Richard M. and Walter Schneider (1977). Controlled and automatic human information processing, II: Perceptual learning, automatic attending, and a general theory. Psychological Review 84: 127-190.
425. Silva-Corvalán, Carmen (1995). Contextual conditions for the interpretation of poder and deber in Spanish. In: J. Bybee, S. Fleischman (eds.), 67-105.
426. Simpson, Greg B. (1994). Context and the processing of ambiguous words. In: M. A. Gernsbacher (ed.), 359-374.
427. Sinha, Chris (1994). Review of Nuyts (1992), 'Aspects of a cognitive-pragmatic theory of language'. Cognitive Linguistics 5: 185-189.
428. Slobin, Dan I. (1991). Learning to think for speaking: Native language, cognition, and rhetorical style. Pragmatics 1: 7-25.
429. Slobin, Dan I. (1996). From 'thought and language' to 'thinking for speaking'. In: J. Gumperz, S. Levinson (eds.), 70-96.
430. Slobin, Dan I. and A. A. Aksu (1982). Tense, aspect, and modality in the use of the Turkish evidential. In: P. Hopper (ed.), Tense-aspect: Between semantics and pragmatics. Amsterdam: Benjamins, 185-200.
431. Sperber, Dan and Deirdre Wilson (1981). Irony and the use-mention distinction. In: P. Cole (ed.), Radical pragmatics. New York: Academic Press, 295-318.
432. Steele, Susan (1975a). Is it possible? Working Papers on Language Universals 18: 35-58.
433. Steele, Susan (1975b). On some factors that affect and effect word order. In: C. Li (ed.), Word order and word order change. Austin: University of Texas Press, 197-268.
434. Steele, Susan (1978). The category AUX as a language universal. In: J. Greenberg (ed.), Universals of human language, vol. 3. Stanford: Stanford University Press, 7-45.
435. Stein, Dieter and Susan Wright (eds., 1995). Subjectivity and subjectivisation. Cambridge: Cambridge University Press.
436. Stephany, Ursula (1986). Modality. In: P. Fletcher, M. Garman (eds.), Language acquisition. Cambridge: Cambridge University Press, 375-400.
437. Stephany, Ursula (1993). Modality in first language acquisition: The state of the art. In: N. Dittmar, A. Reich (eds.), 133-144.
438. Swan, Toril and Olaf Jansen Westvik (eds., 1997). Modality in Germanic languages: Historical and comparative perspectives. Berlin: De Gruyter.
439. Sweetser, Eve (1988). Grammaticalization and semantic bleaching. Berkeley Linguistic Society 14: 389-405.
440. Sweetser, Eve (1990). From etymology to pragmatics. Cambridge: Cambridge University Press.
441. Talmy, Len (1988a). Force dynamics in language and cognition. Cognitive Science 12: 49-100.
442. Talmy, Len (1988b). The relation of grammar to cognition. In: B. Rudzka (ed.), Topics in cognitive linguistics. Amsterdam: Benjamins, 165-205.
443. Talmy, Len (1996). Fictive motion in language and 'ception'. In: P. Bloom, M. Peterson, L. Nadel, M. Garrett (eds.), Language and space. Cambridge, MA: MIT-Press, 211-276.
444. E: Un marqueur d'évidentialité. Langue Française 102: 41-55.
445. Thompson, Sandra A. and Anthony Mulac (1991). A quantitative perspective on the grammaticization of epistemic parentheticals in English. In: E. Traugott, B. Heine (eds.), vol. 2, 313-329.
446. Tiggeler, Eric (1991). Extrapositie en zinsbepalingen. Free University of Amsterdam MA-thesis.
447. Tomasello, Michael (ed., 1998). The new psychology of language. Mahwah: Erlbaum.
448. Tomlin, Russell (1985). Foreground-background information and the syntax of subordination. Text 5: 85-122.
449. Tottie, Gunnel (1983). The missing link? Or, why is there twice as much negation in spoken English as in written English? In: S. Jacobson (ed.), Papers from the second Scandinavian symposium on syntactic variation. Stockholm: Almqvist, 67-74.
450. Traugott, Elizabeth Closs (1980). Meaning-change in the development of grammatical markers. Language Science 2: 44-61.
451. Traugott, Elizabeth Closs (1989). On the rise of epistemic meanings in English: An example of subjectification in semantic change. Language 65: 31-55.
452. Traugott, Elizabeth Closs (1995). Subjectification in grammaticalisation. In: D. Stein, S. Wright (eds.), 31-54.
453. Traugott, Elizabeth Closs (1997). Subjectification and the development of epistemic meaning: The case of promise and threaten. In: T. Swan, O. J. Westvik (eds.), 185-210.
454. Traugott, Elizabeth Closs and Bernd Heine (eds., 1991). Approaches to grammaticalization. Amsterdam: Benjamins.
455. Traugott, Elizabeth Closs and Ekkehard König (1991). The semantics-pragmatics of grammaticalization revisited. In: E. Traugott, B. Heine (eds.), 189-218.
456. Traugott, Elizabeth Closs, Alice ter Meulen, Judy Snitzer Reilly and Charles A. Ferguson (eds., 1986). On conditionals. Cambridge: Cambridge University Press.
457. Urmson, J. O. (1952). Parenthetical verbs. Mind 61: 480-496.
458. Valin, Robert D. Van (1990). Layered syntax in role and reference grammar. In: J. Nuyts et al. (eds.), 193-232.
459. Valin, Robert D. Van (1993). A synopsis of role and reference grammar. In: R. Van Valin (ed.), Advances in role and reference grammar. Amsterdam: Benjamins, 1-164.
460. Valin, Robert D. Van and Randy J. LaPolla (1997). Syntax: Structure, meaning and function. Cambridge: Cambridge University Press.
461. Vandenbosch, Luc (1992). Aspekten van passiefvorming in het Nederlands. University of Antwerp PhD-thesis.
462. Verhagen, Arie (1986). Linguistic theory and the function of word order in Dutch. Dordrecht: Foris.
463. Verschueren, Jef (1985). Review article of Leech (1983), 'Principles of pragmatics', and Levinson (1983), 'Pragmatics'. Journal of Linguistics 21: 459-470.
464. Verschueren, Jef, Jan-Ola Östman and Jan Blommaert (eds., 1995). Handbook of pragmatics: Manual. Amsterdam: Benjamins.
465. Verwijs, E. and J. Verdam (comps., 1885-1929). Middelnederlandsch woordenboek. Den Haag: Nijhoff.
466. Vet, Co (1994). Savoir et croire. Langue Française 102: 56-68.
467. Vries, Jan De (1971). Nederlands etymologisch woordenboek. Leiden: Brill.
468. Vries, M. De and J. A. Te Winkel (comps., 1864-1998). Woordenboek der Nederlandsche taal. Den Haag/Leiden: Nijhoff/Sijthoff.
469. Watts, Richard (1984). An analysis of epistemic possibility and probability. English Studies 65: 129-140.
470. White, Alan R. (1975). Modal thinking. Oxford: Blackwell.
471. Whorf, Benjamin Lee (1956). Language, thought, and reality. Cambridge, MA: MIT-Press.
472. Wijk, N. Van (1912). Franck's etymologisch woordenboek der Nederlandsche taal. Den Haag: Nijhoff.
473. Woodbury, Anthony C. (1986). Interactions of tense and evidentiality: A study of Sherpa and English. In: W. Chafe, J. Nichols (eds.), 188-202.
474. Wunderlich, Dieter (1981). Modalverben im Diskurs und im System. In: I. Rosengren (ed.), 11-53.