Суббота, 20 апреляИнститут «Высшая школа журналистики и массовых коммуникаций» СПбГУ
Shadow

CASE GRAMMAR OF… INTONATION

by V. Phillipov (Sofia, Bulgaria)

Introduction. Оn April 13–15, 1967, at a symposium held at the University of Texas, Austin, on the topic of Universals in Linguistic Theory the late Charles J. Fillmore (1929–2014) launched the Case for Case trend in linguistics [Fillmore 1968], subsequently elaborated it in [Fi1lmore 1977], and the ingenious pun became the starting point of a number of offspring ‘case’ theories that have been exerting stimulating influence in linguistics [cf.: Anderson 1971; Foley, Van Valin 1984; Van Valin 2009; Wierzbicka 1980 et al.]. The trend has won recognition and become widely-used in linguistic research [cf.: The Oxford handbook of case 2009]. 

The three dots in the title are used as a suspension point and also a hesitation pause meant to create a trivial effect of the unexpectedness of intonation as a postmodifier in a position of low transitional probability. Intonation is commonly correlated perceptually with pitch, a vestige from the days of the advent of the sonograph, yet nowadays there has been prevalent evidence that the prosodic features — length, loudness, and pitch — usually work in concert. Their linguistic functions, as Ivić [Ivić 1988: 232] points out, are threefold:

(А) they play the decisive role in the domain of sentence intonation;

(В) they serve to divide utterances into words or syntactic groups, thus facilitating the act of communication;

(С) they may be distinctive on the level of word phonology.

(A) and (B) are mandatory, whereas © is optional. The former point to the widely discussed primacy of the sentence over the word, a position upheld from the ancient Indians to modern times. Thus, compare the categorical statement of Quine on this point [cf.: Wiggins 1971: 25]:

The unit of communication is the sentence and not the word. The point of semantical theory was long obscured by the undeniable primacy, in one respect, of words. Sentences being limitless in number and words limited, we necessarily understand most sentences by construction from antecedently familiar words. Actually there is no conflict here. We can allow the sentences a full monopoly of ‘meaning’ in some sense, without denying that the meaning must be worked out. Then we can say that knowing words is knowing how to work out the meanings of sentences containing them. Dictionary definitions are mere clauses in a recursive definition of the meanings of sentences.

With the growing awareness of the role of intonation in a comprehensive description of (a) language linguists intuitively felt that it was a Janus-like creature looking both at the phonological and the syntactic levels:

An accent language employing relative heights may distinguish old from new or topic from comment, with intonation getting a foothold in the syntax. But the foothold is with one foot; the other one is back there doing its primitive dance [Bolinger 1964].

An excursus, or a concrete example of the problem and the theoretical relevance of the issue. Let me start with an example of seemingly slight prosodic differences of the same segmental substance that can bring about substantial differences in the logical form of the grammatical structure by changing the domain of operation of the three Ts of intonation: tonality, tonicity, and tone. One example will suffice: when I was a school boy I was a chorister, and one of the songs the choir used to perform regularly on various occasions, and I must add — with remarkable ardor, — was “The September Rebels” („Сеп­тем­врий­ци”) whose lyrics were written by the Bulgarian poet Hristo Kurpachev (1911–1943), a guerrilla fighter killed in the resistance movement during World War II and an active member of the antifascist clandestine organizations in the Lovech County of Bulgaria. His poems were popular songs during the Resistance Movement in Bulgaria, and also for the 45 years afterwards. Recently they seem to have fallen into oblivion. 

The musical beats obliterate some of the syntactic / intonation boundaries, and I, together with the other choristers, sang one of the quatrains following the reading in (1): 

(1) ||Отно­во вър­вят септемврийци…||
  ||За все­ки заги­нал герой
  |те мрач­но про­кли­нат убийци:||
  |„Ний пак ви зове­ме на бой!”|

However, it was only recently, while perusing Nikolov [Nikolov 1978], that I realized that I had been completely wrong in my comprehension of the poem. In my mistaken, but possible interpretation, I consistently misread the latter two lines of the quatrain, which originally render the following reading:

(2)||Oтно­во вър­вят септемврийци…||
||За все­ки заги­нал герой
|те мрач­но про­кли­нат:|| „Убий­ци,|
ний пак ви зове­ме на бой!|| [cf.: Ibid.: 295].

The restructuring in terms of tonality of (1) and (2) is the cause of the ambiguity of the song: the lack of an explicit rest in between про­кли­нат [‘ they curse’] and убий­ци [‘murderers’] results in reading the latter as a Direct Object (DO) in (1) instead of an utterance-initial Vocative (Voc) (hence a separate IP) as in (2). The Case Frames of the latter halves of the quatrain in (1) and (2) are different corresponding to a difference in the configuration and the scope of the IPs: in (1) убий­ци [‘murderers’] is the final lexeme of the tonality neutral IP and its initial syllable carries the nuclear fall, whereas in (2) it is a case of a Voc in utterance-initial position (hence a separate IP) followed by an intonationally neutral statement (hence the fall and by default a separate IP). The clear-cut correlations between intonation structure and Case Frames points to their closely determining mutual interface. In this respect, if it is axiomatically accepted that — 

[l] e cas est une catégorie, qui exprime une relation entre deux objets. [Hjelmslev 1935: 96],3

it can be posited that intonation structure gets ‘a foothold in the syntax’ by performing a deterministic grammatical function that gains primacy in the disambiguation of the linguistic unit.4

In media linguistics the issue of disambiguation has always been a relevant one for it determines a one-to-one-reference of the sign. Thus, manipulations of the intonation structure can bring about significant changes in the conceptual structure of the statement. In other words, intonation patterns can be both determined and deterministic, i.e. the intonation pattern is a sign.

The sign character of intonation as a grammatical category. The issue of the sign character of intonation was posed and affirmatively answered in [Phillipov 2009]. This section will focus on the role of intonation as a grammatical category.

Two conversational exchanges (3) and (4) are meant to illustrate a grammatical function of intonation5.

(3) A1: Bill could ask a friend.

B1: \Who? (= Which friend could Bill ask?)

(4) A2: Bill could ask a friend.

B2: /Who? (= Who did you say could ask a friend?)

The structure of both exchanges is syntactically identical (Agent — Experiencer), yet in the deep semantic structure the statements of A are ambiguous. In A1/B1 B1 acts as the typical ‘well-informed and impartial spectator’ and uses the falling tone to inquire about the unmarked focus. In A2/B2 B2 may not have heard the name of the Agent and is, in fact, asking about information intended by A2 as old, but the rising tone in B2’s question brings the figure to the foreground, i. e. the figure is originally unmarked but since the question radically changes it, it becomes marked and the markedness is manifested by relative prominence (i. e. a rise). Indeed, B2 is using an echo-question that is stylistically marked compared to B1. Thus, B1 and B2 restructure in retrospect A1 and A2, respectively, both in terms of the topic / focus articulation and stylistically. The two different signs, being the exponents of two different tones, differ in more than one way. Moreover, since intonation in B2 restructures the case relations in A2, i.e. it empties for the time being the original Experiencer ‘a friend’ from any degree of Communicative Dynamism, ‘Bill’, being the most likely answering move, becomes a Patient; therefore, the resulting interpretation is ‘Bill will undergo the process of asking.’ 

Intonation, thus, persists in displaying a Janus-like behavior all the time: the signifier operates on the concrete level but the interpretant in Peirce’s sense of the term functions on a ‘Chinese box structure’ logic up to the level of discourse in a process of constant rankshifting of units. Language, thus, by means of intonational mechanisms, performs the phenomenon of fluctuation, i. e. “units from a lower structural level can be integrated into a higher level and vice versa, acquiring features of the latter” [Molhova 1981: 47]. Which is the lowest level intonation can display fluctuation on? Indeed, it is the word, but then one-word utterances are not interesting unless used for exemplification or teaching purposes.

Intonation becomes a challenge for analysis when it is the exponent of grammatical relations, and, although such phenomena are usually “around the edge of language,” as [Bolinger 1972] aptly puts it, their effect can be attested in categories which are exponents of core grammatical relations such as case and tense amongst others.6

In German, intonation can function as a grammatical morpheme for the realization of the category of case. In the sentence,

(5) Herr Müller schickte dir das Buch und nicht Anna.7

[lit. Mr Müller sent you this book and not Anna.]

if the nucleus is associated with Müller, then it contrasts with Anna, the former being a subject, and the reading will be ‘Mr Müller sent you this book, not Anna’. If the nucleus is associated with dir (‘to you’), then it contrasts with Anna as indirect object corresponding to the reading ‘Mr Müller sent this book to you, not to Anna.’ On these two readings the only means of resolving the ambiguity caused by case relations is intonation, since the noun Anna, being a proper name, cannot be premodified by an article, the latter being an exponent of the case system in German. If Anna in the latter case is an exponent of a dative case form, then it is only because in the grammatical system of German there is dative as the unmarked exponent of indirect objects of this type. The fact that intonation is a case marker is not a mere competing mechanism, for it competes with a zero segmental marker in this case: rather it is an instance of what Peshkovskij referred to as the “compensatory law” of intonation:

Чем яснее выра­же­но какое-либо син­так­си­че­ское зна­че­ние чисто грам­ма­ти­че­ски­ми сред­ства­ми, тем сла­бее может быть его инто­на­ци­он­ное выра­же­ние (вплоть до пол­но­го исчез­но­ве­ния), и наобо­рот, чем силь­нее инто­на­ци­он­ное выра­же­ние, тем сла­бее может быть грам­ма­ти­че­ское (тоже до пол­но­го исчез­но­ве­ния). [Peshkovskij 1959: 181]8

Lyons suggests an algebraic test of establishing the status of a morpheme by analogy based on the premise that “there is nothing in the definition of the morpheme to imply that it must always be an identifiable segment of the word of which it is a constituent.” [Lyons 1968: 182]

(5) is represented in its two different readings, (5a) and (5b):

(5a) Herr Müller schickte dir das Buch und nicht Anna.

[‘Mr Müller sent you this book, not Anna.’]

(5b) Herr Müller schickte dir das Buch und nicht Anna.9

[‘Mr Müller sent this book to you, not to Anna.’]

(5a) and (5b) are rendered in a simplified formulaic fashion as (6a) and (6b), respectively, whereby the abbreviations of the case marker labels are indicated, being grammatical variables, and the lexical forms they are associated with are left out, being constants:

(6a) S6a = {Nom + Finite Verb + Dat + Acc + Conj + Neg + Nom}

(6b) S6b = {Nom + Finite Verb + Dat + Acc + Conj + Neg + Dat}

Algebraically, (6a) holds true also for (7):

(7) Herr Müller schickte dir das Buch und nicht die Mutter.

[Mr Müller sent you this book, not the mother.]

Likewise, (6b) holds true also for (8):

(8) Herr Müller schickte dir das Buch und nicht der Mutter.

(5a) and (5b) are identical in terms of segmentally bound forms; (7) and (8) differ in terms of grammatical forms: die Mutter in (7) is a Nom Sg, Fem, Common Noun, whereas in (8) der Mutter is a Dat Sg, Fem, Common Noun. The difference between (5a) and (5b) is related to the position of the nucleus, i.e. tonicity, and if the formula of (6a) and (6b) are taken into account, then it must be accepted that the difference in the semantic case structure brings about a difference in intonation of a very deterministic character, i.e. the latter is always a realization of a case marker, and mutatis mutandis a sign.

(5) is not a mere hapax in the grammatical system of German (cf. also e.g. Ich schenke das Bild der Tochter deiner Freundin), neither is German one of the rare languages that a grammatical phenomenon of this kind can occur in. Compare the intonational differences in the following Russian examples and note the meaningful differences they convey:10

(9) Дирек­тор заво­да, |инже­нер, |фран­цуз быст­ро вошли в комнату.

[‘The manager of the plant, the engineer, the Frenchman quickly entered the room.’]

(10) Дирек­тор заво­да, |инже­нер-фран­цуз быст­ро вошли в комнату.

[‘The manager of the plant, the French engineer quickly entered the room.’]

(11) Дирек­тор заво­да, |инже­нер-фран­цуз, |быст­ро вошел в комнату.

[‘The manager of the plant, [who was] a French engineer, quickly entered the room.’]

The construction of a grammar of intonation to account for such cases in a unified approach is an arduous task, for it has to impose various constraints. Note that (5), for once, has to be a paratactic construction whose latter part of the coordination is mandatorily negated and the verb, which is the same for both parts, has to have the [Agent + Experiencer + Object] frame.

Intonation can be the manifestation of temporal systems, too. The following example from Bulgarian is a hard nut to crack for any theoretical model. Yanakiev [Yanakiev 1974] discusses the semantically ambiguous effect due to the different placement of the nuclear accent in utterances having the finite verb form in the pluperfect and followed by an adverbial of time. Thus, compare (12) with (13):

(12) Тоя път той беше излязъл в пет часа.

[‘This time he had left by five o’clock.’]

(13) Тоя път той беше излязъл в пет часа.

[‘This time he had left at five o’clock.’]

Its theoretical importance is, as Yanakiev points out, grounded in the fact that the accent in such cases „сиг­на­ли­зи­ра за нещо пове­че от опре­де­ле­но състо­я­ние на съзна­ни­е­то на гово­ре­щия, носи наблю­да­е­ма „извън­мозъч­на” семан­ти­ка, съоб­ща­ва ни не как­во „иска да под­чер­тае гово­ре­щи­ят в изра­за”, а как­во е обек­тив­но­то отно­ше­ние меж­ду вре­ме­то на едно гла­гол­но дей­ствие и фик­си­ра­на­та чрез няка­к­во обсто­я­тел­стве­но пояс­не­ние ори­ен­та­ция във вре­ме­то.” [Ibid.: 120]11 This is a restricted phenomenon in Bulgarian and it is only pluperfects and the temporarily identical non-evidential forms of the pluperfect (e.g. ‘бил излязъл’ [‘had gone out’] and ‘е бил излязъл’ [‘had gone out’]) associated with them, and also tentatively structures of the past imperfect tense forms of the auxiliary ‘съм’ [‘be’] plus past passive participle forms which display similar semantic behavior:

(14) Тоя път раз­коп­ки­те бяха пре­кра­те­ни през август.

[‘This time the excavations were interrupted in August.’]

(15) Тоя път раз­коп­ки­те бяха пре­кра­те­ни през август.

[‘This time the excavations were interrupted [exactly] in August.’]

Native speakers of Bulgarian will be unanimous as to the difference between (12) and (13), but will not commit themselves in a likely manner as to the difference between (14) and (15): in the latter case the native speaker usually expects wider temporal context or makes inferences about the temporal context where the structures are used. This is yet another proof of the fuzzy nature and the overall complexity of intonational phenomena. One thing that Yanakiev is definite about is that in such cases the nuclear accent functions as a morpheme. As he states explicitly, „[o]собеност на тези мор­фе­ми е, че за тях се сиг­на­ли­зи­ра супра­сег­мент­но…” [Ibid.].12 In conclusion to this section, it can be posited that intonation is a multidirectional vector acting on a multilevel scale and based deep in our innate facultas signatrix, or, as Kant [Kant 1983] put it more than two centuries ago: “Das Vermӧgen der Erkenntnis des Gegenwӓrtigen als Mittel der Verknüpfung der Vorstellung des Vorhergesehenen mit der des Verhangenen des ist das Bezeichnungsvermӧgen. — Die Handlung des Gemüts diese Verknüpfung zu bewirken ist die Beziechnung (signatio), die auch das Signalieren genannt wird, von der nun der grӧβere Grad die Auszechnung genannt wird.” [Ibid.: 117]13 At this stage one is reminded of Nikolaeva’s felicitous term „скры­тая память язы­ка” [‘the hidden memory of language’], and that „[в] этом слу­чае мож­но ска­зать, что язык не столь­ко пом­нит, сколь­ко зна­ет, что он обя­зан выра­зить то или иное содер­жа­тель­ное про­ти­во­по­став­ле­ние.” [Nikolaeva 2013: 34] The efficiency of the sign lies in the all-embracing organizational power it possesses, wherever its provenance lies in. For all intents and purposes, as Kant somewhat tongue in cheek suggests, thinking is the language of the stomach, for at least that is what the Indians of Tahiti believed: “… die Indianer auf Otaheite nennen das Denken: die Sprache im Rauch” [Kant 1983: 119–120].

An increase of the delicacy of analysis. This section aims at increasing the delicacy of analysis by establishing a more deterministic syntax/ semantics interface that intonation, viewed sensu largo as the function of all prosodic features on sentence-level conveying linguistic and pragmatic meaning, is instrumental in setting up. The term ‘delicacy’, belonging to the Hallidayan tradition, is used to “refer to one of the scales of analysis which interrelates the categories of the theory, viz the dimension which recognizes increasing depth of detail.” [Crystal 1980: 104] The approach propounded is presented as an alternative to an already existing one of Functional Sentence Perspective (FSP) as expounded in Firbas [Firbas 1992], a seminal work representative of the best of the Prague School tradition. The focus of analysis will be an article by Firbas [Firbas 1999].

To start with, two principles, (P1) and (P2), will be taken for granted:

(P1)Talk is cheap (cf. [Haiman 1998]);

(P2) Initially, language was used by people when they had to “do” something.

(P1) and (P2) can be taken as axioms.

The Praguians were the first to set up a comprehensive approach to the interaction between the grammatical structure of the sentence and its semantic structure, and by extension the FSP organization of the utterance on a synchronic level. Three levels are distinguished:

the level of the grammatical structure of the sentence;

the level of the semantic structure of the sentence;

the level of the organization of the utterance. [Daneš 1966: 225]

My aim is to enhance the power of (i) and (ii) by establishing a more deterministic interface by making using a Role and Reference Grammar (RRG) [cf.: Foley, Van Valin 1984; Van Valin 2009 et al.], an offspring of Case Grammar, and, thus making (iii) a derivative of (i) and (ii). 

Going back to [Firbas 1999], (16a) and (17a) illustrate a Firbassian approach, and (16b) and (17b) present a RRG analysis of the ambiguous sentence (Dogs must be carried on the escalator.). The words in capitals indicate the foci.

(16) Dogsmust be CARRIEDon the escalator.
B(earer of Quality)Q(uality)Set(ting)
U(ndergoer), P(rivileged) S(yntactic) (Argument), Marked PivotActivity; verb DO (x, [carry’ (dogs)]; Fo(cus)NMR (non-macrorole argument), Loc(ative)
(17) DOGSmust be carriedon the escalator.
Ph(enomenon)Pr(esentation of Phenomenon)Set
U; PSA; FoActivity verb DO (x [carry’ (dogs)]NMR; Loc

RRG posits two generalized semantic relations, termed macrorole arguments (MR’s), between a predicate and its arguments: an A(ctor) and an U(ndergoer). The former (by people) in (16) is deleted as a result of the passive transformation rule. These two relations constitute an interface between syntactic relations such as subject and semantic relations such as case roles or thematic relations. In a simple active clause the MR’s, traditionally called the ‘subject’ and the ‘object’, or jointly the ‘core’ as it is referred to in RRG, together with the predicate, or the ‘nucleus’, respectively, can be surrounded by an outermost layer, or ‘periphery’, whose exponents are NMR’s, the latter usually being optional since the latter do not form part of the predicate’s valency. Due to postulate (P1), when language structure itself has no other formal means left, language employs intonational means to manifest exponents of semantic case marking by resorting to the “hidden memory of language”. In the case of (16) language ‘remembers’ that when a marked pivot has to be marked as being focal, the ultimate default semantic mechanism language can offer is intonation. With respect to language evolution intonation was universally grammaticalized first compared to the morphological exponents of case marking, at the times when it was still “doing its primitive dance”. This is why it will be always impossible, as far as intonation is concerned, “to separate the linguistically arbitrary from the psychologically expressive.” [Bolinger 1964: 844]

Thus, intonation, by virtue of the fact that it always verges on gesture, the latter always being relational in human communication, can and, in the remote past, did serve primordially as a semantic case marker. Such a hypothesis awaits its verification for the ergative type of languages, and the content typology approach in this direction is promising to bring its own share of surprises. Its validation will come as yet another proof of Karcevskij’s thesis that “[s]i intellectualisée et appauvrie soit-elle, l’intonation fait néanmoins partie intégrante du mécanisme linguistique.”14 [Karcevskij 1964: 211]

Last but not least, an increase of the delicacy of analysis of the kind discussed above will be a useful tool in the theoretical inventory of media linguistics with special reference to texts rendered in the spoken medium of language. In a seminal article on the status of media linguistics Gajda sketches out the following three factors as being characteristic of what he terms „инте­лек­ту­аль­ная аура в совре­мен­ных гума­ни­тар­ных нау­ках” (the intellectual aura in modern humanities) [Gajda 2015: 18]:

Reality is not always a manifestation of moderate realism; in some instances it is manifested by set construals, i.e. a process of moderate conservatism.

Knowledge of the world is contingent on the countless contexts it can be generated by, i.e. contextualism.

Human knowledge is characterized by the availability of a multitude of perspectives, i.e. epistemological pluralism.

Whereas (a) and (b) are mandatory features of other approaches to the domain of the humanities (e.g. positivism), © seems to stand out as the ‘distinctive feature’ in the currently dominating scientific paradigm. The methodology put forward in the present article establishes © as its scientific dominant. 

Envoi. A lonely voice in the late fifties of the 20th century, which tried to break the shackles of the linear (= one directional on a Euclidean plane) nature of the sign, but remained unnoticed in the West, came from the Russian Anglist A. I. Smirnitskij who aptly points out that if questions such as Has he come? and Is the book on the table? are realizations of one and the same IP, irrespective of the fact that their sound content is completely dissimilar, then, it is better to talk of “bilinearity” [‘двух­ли­ней­ность’], rather than “linearity”. [Smirnitsky 1957: 17] The idea does not quite come to the point in terms of present-day linguistic achievements (cf. multi-linear phonology), but it did point in those days, though tragically ahead of its time, to the right direction, at least structurally so: the idea of mapping (two) lines was already in the air. 

Traditionally, linguistics has been resorting to a level type of analysis (phonology — morphology — syntax) in a monosystemic way. What is needed is a polysystemic approach whereby „за осно­ву берут­ся семан­ти­че­ские кате­го­рии грам­ма­ти­ки и семан­ти­че­ские функ­ции, объ­еди­ня­ю­щие раз­но­уров­не­ные язы­ко­вые сред­ства. Линг­ви­сти­че­ский ана­лиз при таком под­хо­де направ­лен на изу­че­ние вза­и­мо­дей­ствия эле­мен­тов раз­ных уров­ней на функ­ци­о­наль­ной осно­ве.” [Bondarko 1987: 7]15 Intonation perfectly lends to a polysystemic approach for it is the wandering Jew in language: it goes through all levels but never anchors permanently at any of them. Its spheres of intensity depend on the typological specificities of the specific language. In terms of substance intonation is related to phonology; functionally it acts on units of all levels, and communicatively is manifested in speech. It is a suprasystem — to use a term of Pencheva [Pencheva 1983]16 — and as such is like semantics, the latter also embracing all levels but in a qualitatively different way. The systems of intonation interact with the sign as a whole whereas semantics is merely one side of it, namely the plane of content. Intonation is also the archetypal wanderer on the time dimension of language, and like any temporal phenomenon is subject to changes on the temporal axis: like a summer storm it can get stronger but it can also abate, i.e. it has its peaks and troughs. And still in consonance with time, it is always a continuum. It is thus obvious that the dyadic relation of an aliquid stat pro aliquo definition of the sign cannot account for all the idiosyncracies of the eternal wanderer. 

1 The September rebels are on the march again…
And for every hero who perished
Gloomily the murderers they cursed:
“We call you to fight again!”

All translations are mine. The symbols (|) and (||) represent the boundaries between Intonation(al) Phrases (IP’s). The last term is an exponent of the tendency for the IP to coincide in extent with the clause, hence the IP will be regarded in such cases as “neutral in tonality”. On this point, cf. [Halliday 1967: 18–19] although the latter uses ‘tone group’ for ‘IP’. I use ‘IP’ in line with [Wells 2006], the latter following the modern American tendency [cf.: Pierrehumbert 1980].

2 The September rebels are on the march again…
And for every hero who perished
Gloomily they cursed : “Murderers,
We call you to fight again!”

3 “Case is a category that expresses a relation between two objects.”

4 In structuralist terms the Vocative, termed an ‘outlier case’ [cf.: Daniel & Spencer 2009: 626], is usually considered ‘extragrammatical’, because it is not an exponent of a grammatical relation, i.e. there are no grammatical oppositions it can participate in. In terms of Saussure’s understanding of ‘opposition’ it should be part of a paradigmatic relation, and in (2) убий­ци [‘murderers’] is a unit of a different discoursal move compared to (1). The difference between (1) and (2), however, brings to the fore a difference of contrast, i.e. a syntagmatic one, which on the level of discourse is irrelevant, still for the language user, when sung (or read without the punctuation marks) it differentiates meaningful utterances. Such a consideration leaves the case status of the Voc open and unresolved. 

5 The examples are from [Wells 2006: 17].

6 See [Jun, Fletcher 2014: 502]: “Intonation can also mark a change in word order (e. g. Georgian,….) or function as a morpheme (e. g. the absolutive case marker in Samoan,…).” 

7 The example was initially discussed in [Phillipov 2009; 2012]. 

8 “The clearer a syntactic meaning is realized by purely grammatical means, the weaker will be its intonational expression (till its complete disappearance), and vice versa, the stronger the intonational exponent, the weaker will be the grammatical one (also till its complete disappearance).”

9 The bold types indicate nuclear accent in the lexeme.

10 Examples (9), (10) and (11) are from Gvozdev [Gvozdev 1963: 166].

11 “… it signals more than a mere state of awareness of the speaker’s state of consciousness; it conveys observable, “outside-the-brain” semantics; it tells us not “where in the structure the speaker wants to lay its emphasis”, but what the objective correlation between the time of a verbal activity and the orientation in time fixed by some adverbial of time.”

12 “What is particular about these morphemes is that they are signaled suprasegmentally…”

13 “The ability to recognize the present as the means for connecting ideas of foreseen events with those of past events is the power of using signs. — The essential activity of making this connection is signifying (signatio), which is also called signaling. If it is present in a higher degree, it is also called characterizing.”

14 “… no matter how intellectualized and impoverished it is, intonation is still an integral part of the linguistic mechanism.” 

15 “the foundations are formed by semantic categories of grammar and semantic functions grouping linguistic units from different levels. The linguistic analysis on such an approach aims at the study of the interaction of elements from different levels on a functional basis.”

16 Pencheva uses the term with reference to the roleplayed by word-formation, another wandering Jew in linguistic structure.

© Phillipov V., 2017