Lidija Iordanskaja and Igor Mel’čukcoord<strong>in</strong>ates of P), or on the clause P; <strong>in</strong> this case, we speak of illocutive Specifiers. S<strong>in</strong>ce anIPVE constitutes a k<strong>in</strong>d of a comment, by the Speaker, on the clause P (Bonami and Godard2007: 259 and K&P 2011), it is an illocutive Specifier. The comment conveyed by an IPVEcan bear on:— The epistemological status of P, <strong>in</strong>clud<strong>in</strong>g the <strong>in</strong>dication of the Speaker’s source of the <strong>in</strong>formation« P » (‘I believe’, ‘I swear it’, ‘<strong>de</strong>clared the m<strong>in</strong>ister’, ‘as our calculations show’).— The subjective attitu<strong>de</strong> of the Speaker or anybody else towards the content of P (‘I amafraid’, ‘I regret’, ‘the government hopes’).— The l<strong>in</strong>guistic form of P (‘as say the Spaniards’).Locutionality. From the viewpo<strong>in</strong>t of their locutionality, IPVEs are signalatives (Mel’čuk2001: 245ff, 354ff)—more precisely, syntactic signalatives: their signalative character is expressedby a parenthetical syntactic construction. A signalative is a mean<strong>in</strong>g ‘σ’ reflect<strong>in</strong>g apsychological state of the Speaker or a rhetorical action by him such that he verbalizes it bysignal<strong>in</strong>g (rather than by communicat<strong>in</strong>g): a prototypical signalative expression does not allowfor negation, 3 <strong>in</strong>terrogation or free modification; it never constitutes an assertion <strong>in</strong> logicalsense. Thus, the IPVEs <strong>in</strong> sentences (1) present a rhetorical action by the Speaker—namely,the <strong>in</strong>troduction of an <strong>in</strong>ci<strong>de</strong>ntal comment concern<strong>in</strong>g P (<strong>in</strong> this case, the signaled mean<strong>in</strong>g is‘my source of the <strong>in</strong>formation «P» is E’).To make the notion of signalative clearer, here are examples of signalatives that are notIPVEs.• First, there are lexical signalatives (= lexical units stored as such <strong>in</strong> the lexicon):<strong>in</strong>terjections as Wow! or Phew!; parenthetical adverbials such as unfortunately, to mysense or of course; connector adverbials such as <strong>in</strong> fact or for <strong>in</strong>stance; rhetoricalconjunctions such as although or s<strong>in</strong>ce; etc.• Second, there are morphological signalatives, for <strong>in</strong>stance, the imperative.• F<strong>in</strong>ally, there are syntactic signalatives, such as the <strong>Russian</strong> construction «V INF -PAR-TICLE “TO” X V FIN » (Čitat´-to on čital, da... lit. ‘To.read-“to” he has.read, but…’ ≈‘Although he has read [it], but…’), which signals the skepticism of the Speaker withrespect to X’s action V.At this po<strong>in</strong>t, two important remarks seem to be appropriate.1. Note that an illocutive Specifier is not necessarily Signaled—that is, <strong>in</strong> our case, it doesnot to be expressed as an IPVE. Thus, <strong>in</strong> I believe that the situation is <strong>de</strong>teriorat<strong>in</strong>g the boldfacedmatrix clause can be, <strong>in</strong> a particular context (for <strong>in</strong>stance, as an answer to the questionWhat is happen<strong>in</strong>g there?), an illocutive Specifier without be<strong>in</strong>g Signaled: it is not an IPVE.2. The communicative status «Specifier + Signaled» of a mean<strong>in</strong>g results <strong>in</strong> its weak communicativevalue; many researchers (e.g., Z&P 1987: 84 and K&P 2011) consi<strong>de</strong>r thisproperty as essential for IPVEs. A weak communicative value also characterizes Backgroun<strong>de</strong><strong>de</strong>xpressions, such as the boldfaced clause <strong>in</strong> My friends (who live <strong>in</strong> Canada) likeski<strong>in</strong>g. But this is another communicative opposition, irrelevant <strong>in</strong> the context of this talk,—Perspective (Mel’čuk 2001: 198ff); an IPVE can be Neutral, as <strong>in</strong> (7a), or Backgroun<strong>de</strong>d, as<strong>in</strong> (7b):3 More precisely, negation cannot bear on the central component of a signaled mean<strong>in</strong>g. Thus, <strong>in</strong> the case of theimperative (which is a morphological signalative)—e.g., Don’t say this!—the central component of animperative mean<strong>in</strong>g ‘I want you to …’ is not negated.123
<strong>Illocutive</strong> <strong>Parenthetical</strong> <strong>Verbs</strong> <strong>in</strong> <strong>Russian</strong>(7) a. Položenie, kak sčitajut vse, uxudšaetsja‘The situation, as everybody believes, is <strong>de</strong>teriorat<strong>in</strong>g’.b. Položenie (kak sčitajut vse) uxudšaetsja.Therefore, an IPVE cannot be <strong>de</strong>f<strong>in</strong>ed simply as be<strong>in</strong>g communicatively Backgroun<strong>de</strong>d.The communicative role of an IPVE is manifested ma<strong>in</strong>ly by its prosody: an IPVE allows(or requires) pauses at its boundaries, carries—<strong>in</strong> a neutral context—a flat <strong>in</strong>tonationalcontour and cannot have sentential stress (Z&P 1987: 81-82; Bonami and Godard 2007: 262speak of “<strong>in</strong>ci<strong>de</strong>ntal prosody”). It is prosody that dist<strong>in</strong>guishes a verbal IPVE <strong>in</strong> the <strong>in</strong>itialposition <strong>in</strong> the sentence (8a) from its non-parenthetical counterpart with the ellipsis of theconjunction ČTO ‘that’:(8) a. (i) Napom<strong>in</strong>aju, (|) Maša bol´nÁ ‘[I] rem<strong>in</strong>d, Masha is ill’. ≡3(ii) Maša, | napom<strong>in</strong>aju, | bol´nÁ ‘Masha, [I] rem<strong>in</strong>d, is ill’. ≡b. (i) Napom<strong>in</strong>Áju, || Maša bol´nÁ ‘[I] rem<strong>in</strong>d, Masha is ill’. ≡(ii) Napom<strong>in</strong>Áju, čto Maša bol´nÁ ‘[I] rem<strong>in</strong>d that Masha is ill’.In (8a), napom<strong>in</strong>aju ‘[I] rem<strong>in</strong>d’ is an IPVE; <strong>in</strong> (8b), napom<strong>in</strong>aju constitutes the matrixclause that governs its completive—asyn<strong>de</strong>tically <strong>in</strong> (8b-i) and by means of ČTO (8b-ii).2.2 The Syntactic Role of an IPVEAt the Deep-syntactic level, an IPVE <strong>de</strong>pends on the head of the clause P by the <strong>de</strong>epsyntacticrelation APPEND, which represents all k<strong>in</strong>d of “extrastructural” constructions,manifest<strong>in</strong>g weak subord<strong>in</strong>ation, such as sentence adverbs, parenthetical expressions,addresses, <strong>in</strong>terjections, prolepses, etc. This type of subord<strong>in</strong>ation is opposed to strong subord<strong>in</strong>ation—thatis, actants and modifiers/circumstantials.Three Syntactic Constructions for IPVEs <strong>in</strong> <strong>Russian</strong>The constructions un<strong>de</strong>r analysis will be illustrated by the verb SČITAT´ ‘believe’.3.1 Non-<strong>Parenthetical</strong> Use of SČITAT´Consi<strong>de</strong>r first a non-parenthetical use of the verb SČITAT´:(9) a. Ja sčitaju, čto položenie uxudšaetsja ‘I believe that the situation is <strong>de</strong>teriorat<strong>in</strong>g’.Here is its semantic representation [= SemR]:b.ThemeRheme SEM -1SEM -1‘<strong>de</strong>teriorat<strong>in</strong>g’‘believe’ 2 RhemeSEM-211 1‘situation’‘now ‘I’ThemeSEM-2In a communicative area, un<strong>de</strong>rl<strong>in</strong><strong>in</strong>g <strong>in</strong>dicates the communicatively dom<strong>in</strong>ant no<strong>de</strong>, i.e., the semanteme thatrepresents the m<strong>in</strong>imal paraphrase of the area’s mean<strong>in</strong>g. The semanteme ‘now’ is an abbreviation thatenco<strong>de</strong>s the present <strong>in</strong>dicative of the verb.124