vocatives and other direct address forms: a contrastive study

In this paper I analyze Russian direct address forms, both the distinct truncated vocative and nominative-case direct address forms. I contrast the formal and functional restrictions on the truncated vocative with vocatives in other languages (e.g. Czech and Polish), and I compare the interpolation of Russian direct address forms in an utterance to the situation in English. While similarities are found both in the form and the usage of Russian direct address formswith those in other languages, the prosodic and syntactic constraints in English are considerably stronger than in Russian, which means that the punctuating function of direct address forms is considerably more flexible in Russian than in English. «What do you mean?» «I’m seriously thinking I may resign, Jim.» The fact that he used my name seemed almost as important as the statement that preceded it. Was he saying one thing or two? Don DeLillo, The Names. [1] introduct ion The aimof this paper1 is to compare certain aspects of Russian direct address form usagewith that in other languages, especially English, Czech, and Polish. By direct address form I mean any expression used to attract or maintain the addressee’s attention, as in (1), as opposed to other, non-address usages (e.g. as arguments), as in (2): (1) Mr. Smith, could you tell us about your trip to Washington? [1] I wish to thank the two anonymous reviewers for their detailed comments on an earlier draft of this article. [212] lillian a. parrott (2) a. Mr. Smith went to Washington. b. I saw Mr. Smith on the train. The term vocative will, in contrast, be used in a restricted sense to refer only to direct address forms that are formally distinct from the nominative.2 In Russian there are two types of distinct forms: the truncated vocative (the so-called ”neovocative”),3 as in (3), which is the only productive vocative in modern Russian; and the historical Slavic vocative, as in (4), which is found in a few frozen forms used mainly as invocations or interjections:4 (3) мам! Mom-VOC, Тань!, Tanja-VOC, Саш!, Saša-VOC, Никит!, Nikita-VOC, теть Aunt-VOC Ань!, Anja-VOC, ребят! kids-VOC (4) боже!, god-VOC, господи! lord-VOC In the presentation below, we will consider truncated vocatives (as in (3)) and non-truncated direct address forms, but the historical Slavic vocative remnants in Russian, as in (4), will not be discussed further here since they only exist as frozen forms. [2] the truncated vocat ive in russ ian [2.1] Formal peculiarities of the Russian truncated vocative The Russian truncated vocative presents certain peculiarities that set it apart from other forms (from Parrott (1993, 1995); see Daniel’ (2009) for a more recent discussion of these factors): (a) The truncated vocative is formed on personal names and kinship terms having a penultimate-stressed nominative in -a (e.g. Petrúšk! [<Petrúška] but not [2] I am following Daniel & Spencer (2009) in reserving the term ”vocative” for forms that are distinct from the nominative, but this distinction between vocative and direct address form is by no means widely observed (Daniel and Spencer use the term form of address instead of direct address form). Linguistsworking on direct address forms in languages where there is no distinct vocative form generally prefer the term ”vocative” over ”address form” or ”form of address”, since the latter terms are commonly used to speak of the choice of expression used to refer to a person, rather than forms used specifically to address a person directly. But since Russian can use both distinct vocative forms (as in (3)) and regular nominativecase direct address forms, some terminological distinction needs to be made between the two types, and I hope that choosing the term direct address form will remove some of the potential ambiguity. [3] As shown in Parrott (1993) and, more recently, Daniel’ (2009), the so-called neo-vocative is not so new; it is reported from the mid-19th century (see Obnorskij (1925)) and could be considerably older even though not attested. [4] I use the exclamation point to mark vocatives and other direct address forms used in isolation even though they need not be uttered as exclamations. OSLa volume 2(1), 2010 vocatives in contrast [213] *Ivánušk!5 [<Ivánuška]), i.e. mostly singular a-declension nouns,6 but the vocative forms rebjat! [<rebjata] ”kids!/guys!” and devčat! [<devčata] ”girls” are irregular plurals. (b) If a non-diminutive familiar form exists for a first name (e.g. Jura < Jurij), truncated vocative are not normally formed from the full name: Inn! [<Inna] and Nikit! [<Nikita] are acceptable because they have no familiar forms, only diminutives, but ???-*Ol’g! [<Ol’ga] and ???-*Ann! [<Anna]7 are usually not possible because they have familiar forms Olja and Anja, which can undergo vocative truncation: Ol’!, An’!. This restriction appears to be weakening, however, at least for some names; Daniel’ (2009, 233-234) cites some (mostly, but not exclusively, recent) examples from the Russian National Corpus of truncated vocatives from full first names that do have familiar forms (e.g. Svetlan!), but their usage remains marginal. (c) The truncated vocative does not produce vowel-zero alternations (e.g. Jurk!), whereas elsewhere in the system such alternations are required (cf. U nas v detskom sadu neskol’ko Jurok/*Jurk). (d) The truncated vocative does not cause mandatory devoicing of consonants word finally (Nad’! [nad’]/[nat’]), whereas elsewhere in the system such devoicing is mandatory (cf. V klasse bylo mnogo Nad’ [nat’]/*[nad’]). In Parrott (1993, 78) it was suggested that the peculiarities given in (c) and (d) could be accounted for by positing a voiceless or devoiced vowel as the ending, which leaves the underlying structure intact, in which case the term ”truncated” is something of a misnomer. Daniel & Spencer (2009, 629) consider that the Russian vocative is an example of phonological truncation (as opposed to morphological truncation). [2.2] Functional peculiarities of the Russian truncated vocative The Russian vocative also presents certain peculiarities in its usage, as compared to vocatives in other languages (from Parrott (1993) and Parrott (1995)). [5] My Russian examples have been evaluated by a number of different speakers: firstly by a group of speakers who grew up in the Soviet Union (1 female from Leningrad b. circa 1930, 1 male fromMoscow b. circa 1940, 1 female from Sotchi b. circa 1950, 2 females from Moscow b. circa 1960, 1 female from Leningrad b. circa 1960), and secondly by a group of speakers who came of age in the post-Soviet period (1 male and 1 female from Moscow b. circa 1980, 1 female from the Petersburg area b. circa c. 1985). [6] Russian short first names and diminutives generally belong to this class, and thus most informal or familiar names are open to vocative truncation. See Nesset (2001) on the notion of familiarity associated with the a-declension. [7] Ann exists as the foreign first name Ann(e), and Ol’g exists as a last name. OSLa volume 2(1), 2010 [214] lillian a. parrott (a) The truncated vocative is only used to address people, or sometimes superior animals, e.g. sobak! ”doggie!”, but not inanimates.8 (b) The truncated vocative is optional, and is generally restricted to a relatively informal setting, with a relatively close interlocutor relationship (signalled in part by the usage of familiar first names and ty-address, although these are not absolute indicators);9 formality or any other kind of (momentary) distancing in the interlocutor relationship or the subject of discourse (respect, solemnity, anger, aggressiveness, etc.) renders its usage impossible or unlikely. The notions in (b) are of course very fuzzy and the boundaries vary greatly from speaker to speaker and situation to situation, and there are also trends according to generation, region, and social class. Among my informants, the most striking difference has to do with generation: older informants judged (5a) as impossible or nearly so, whereas younger informants, those that came of age in the postSoviet period, especially in urban centers, were much more lenient in their judgments. (5) a. ?-??? Здравствуйте, Hello, Марь Marija-VOC Иванн! Ivanovna-VOC! (made-up example) student to distinguished professor (whom s/he does not know well); in a formal setting b. Здрасьте, Hi, Марь Marija-VOC Иванн! Ivanovna-VOC! (made-up example) to a neighbor in a communal apartment Momentary distancing can occur in otherwise close interlocutor relationships, due to the formality, solemnity, or seriousness of the particular situation and discourse topic, or due to attitudinal factors such as anger or other intense emotions, in which case the truncated vocative is also unnatural, as shown in the examples below. [8] See Daniel’ (2009) for further discussion of some of the restrictions on the types of address forms that can undergo vocative truncation in Russian; invectives, for example, are generally excluded. [9] See Yokoyama (1994) on the truncated vocative as iconic for a close interlocutor relationship. OSLa volume 2(1), 2010 vocatives in contrast [215] (6) #10Витя/???-*Вить, сегодня умер твой отец.11 (note the formal word order)12 #Vitja-NOM/???-*Vitja-VOC, your father died today. (7) uttered in anger:13 a. Миша Miša-NOM / / ???-*Mиш, ???-*Miš-VOC, я I убью could kill тебя! you! b. Я I убью could kill тебя, you, Миша Miša-NOM / / ???-*Mиш! ???-*Miš-VOC! Note that there are similar pragmatic restrictions on the usage of the English attention-getting particle hey, e.g. Betsy/??Bets/???-*Hey Betsy/???-*Hey Bets, I could kill you!, although additional factors come into play as well. We will return to the pragmatic constraints on the truncated vocative further below. [3] uses of truncat ion in other languages [3.1] Elsewhere in Slavic Besides Russian, truncated vocatives are marginal in Slavic, and attested examples are hard to come by. Still, Anstatt (2003); Anstatt & Gut (2008) cite trunc

In this paper I analyze Russian direct address forms, both the distinct truncated vocative and nominative-case direct address forms. I contrast the formal and functional restrictions on the truncated vocative with vocatives in other languages (e.g. Czech and Polish), and I compare the interpolation of Russian direct address forms in an utterance to the situation in English. While similarities are found both in the form and the usage of Russian direct address forms with those in other languages, the prosodic and syntactic constraints in English are considerably stronger than in Russian, which means that the punctuating function of direct address forms is considerably more flexible in Russian than in English.
«What do you mean?» «I'm seriously thinking I may resign, Jim.» The fact that he used my name seemed almost as important as the statement that preceded it. Was he saying one thing or two?
Don DeLillo, The Names.
[1] i n t r o d u c t i o n The aim of this paper 1 is to compare certain aspects of Russian direct address form usage with that in other languages, especially English, Czech, and Polish. By direct address form I mean any expression used to attract or maintain the addressee's attention, as in (1), as opposed to other, non-address usages (e.g. as arguments), as in (2): (1) Mr. Smith, could you tell us about your trip to Washington?
[1] I wish to thank the two anonymous reviewers for their detailed comments on an earlier draft of this article.
(2) a. Mr. Smith went to Washington. b. I saw Mr. Smith on the train.
The term vocative will, in contrast, be used in a restricted sense to refer only to direct address forms that are formally distinct from the nominative. 2 In Russian there are two types of distinct forms: the truncated vocative (the so-called "neovocative"), 3 as in (3), which is the only productive vocative in modern Russian; and the historical Slavic vocative, as in (4), which is found in a few frozen forms used mainly as invocations or interjections: 4 ( In the presentation below, we will consider truncated vocatives (as in (3)) and non-truncated direct address forms, but the historical Slavic vocative remnants in Russian, as in (4), will not be discussed further here since they only exist as frozen forms.
[2] t h e t r u n c at e d vo c at i v e i n r u s s i a n [2.1] Formal peculiarities of the Russian truncated vocative The Russian truncated vocative presents certain peculiarities that set it apart from other forms (from Parrott (1993Parrott ( , 1995; see  for a more recent discussion of these factors): (a) The truncated vocative is formed on personal names and kinship terms having a penultimate-stressed nominative in -a (e.g. Petrúšk! [<Petrúška] but not [2] I am following Daniel & Spencer (2009) in reserving the term "vocative" for forms that are distinct from the nominative, but this distinction between vocative and direct address form is by no means widely observed (Daniel and Spencer use the term form of address instead of direct address form). Linguists working on direct address forms in languages where there is no distinct vocative form generally prefer the term "vocative" over "address form" or "form of address", since the latter terms are commonly used to speak of the choice of expression used to refer to a person, rather than forms used specifically to address a person directly. But since Russian can use both distinct vocative forms (as in (3)) and regular nominativecase direct address forms, some terminological distinction needs to be made between the two types, and I hope that choosing the term direct address form will remove some of the potential ambiguity. [3] As shown in Parrott (1993) and, more recently, , the so-called neo-vocative is not so new; it is reported from the mid-19th century (see Obnorskij (1925)) and could be considerably older even though not attested. [4] I use the exclamation point to mark vocatives and other direct address forms used in isolation even though they need not be uttered as exclamations. In Parrott (1993, 78) it was suggested that the peculiarities given in (c) and (d) could be accounted for by positing a voiceless or devoiced vowel as the ending, which leaves the underlying structure intact, in which case the term "truncated" is something of a misnomer. Daniel & Spencer (2009, 629) consider that the Russian vocative is an example of phonological truncation (as opposed to morphological truncation).

[2.2] Functional peculiarities of the Russian truncated vocative
The Russian vocative also presents certain peculiarities in its usage, as compared to vocatives in other languages (from Parrott (1993) and Parrott (1995)). Ann exists as the foreign first name Ann(e), and Ol'g exists as a last name.
[214] lillian a. parrott (a) The truncated vocative is only used to address people, or sometimes superior animals, e.g. sobak! "doggie!", but not inanimates. 8 (b) The truncated vocative is optional, and is generally restricted to a relatively informal setting, with a relatively close interlocutor relationship (signalled in part by the usage of familiar first names and ty-address, although these are not absolute indicators); 9 formality or any other kind of (momentary) distancing in the interlocutor relationship or the subject of discourse (respect, solemnity, anger, aggressiveness, etc.) renders its usage impossible or unlikely.
The notions in (b) are of course very fuzzy and the boundaries vary greatly from speaker to speaker and situation to situation, and there are also trends according to generation, region, and social class. Among my informants, the most striking difference has to do with generation: older informants judged (5a) as impossible or nearly so, whereas younger informants, those that came of age in the post-Soviet period, especially in urban centers, were much more lenient in their judgments. We will return to the pragmatic constraints on the truncated vocative further below.
[3] u s e s o f t r u n c at i o n i n o t h e r l a n g u a g e s

[3.1] Elsewhere in Slavic
Besides Russian, truncated vocatives are marginal in Slavic, and attested examples are hard to come by. Still, Anstatt (2003); Anstatt & Gut (2008) cite truncated forms in Ukrainian and Polish, and Stankiewicz (1977Stankiewicz ( /1986 gives truncated forms from Bulgarian and Belarusian dialects, in addition to Ukrainian dialectal forms where the final syllable is truncated. Stankiewicz also gives examples of truncated imperatives as expressive variants in Bulgarian, Croatian/Serbian, and Ukrainian. 14

[3.2] In other languages and dialects
Vocative truncation is in fact fairly widespread in the languages of the world. In European languages (e.g. Greek, Baltic) 15 truncation of a final consonant (with or without reduction or other alteration of the preceding vowel) is common in vocative formation, as is truncation of entire syllables, which often occurs in im- [10] I use the pound key (#) to indicate a discourse-initial utterance where this is important for the interpretation under consideration.
If the context is slightly changed to render it more informal, with the contact already established (i.e. where it is no longer attention-getting), the usage of the truncated form becomes possible, e.g. Знаешь, Вить, сегодня умер твой отец, 'You know, Vitja-VOC, your father died today.' I wish to thank Elizaveta Khachatourian for this observation. [12] See Yokoyama (1986) on the formality of this word order with type I intonation, and Yokoyama (1993) on the svoj-čužoj distinction. [13] But with different word order, the truncated form becomes possible (and is attested on the internet), when not uttered in real anger: Mиш, я тебя убью! 'Misha-VOC, I could kill you!' [14] On the affinity between the vocative and the imperative, see, for example, Jakobson (1960Jakobson ( /1981, Winter (1969), Khrakovskij & Volodin (1986), and also Parrott (1993) [15] See Winter (1969) for a discussion of vocative formation in the history of Indo-European. [216] lillian a. parrott peratives as well (e.g. Sardinian and Romanian dialects). 16 In Russian, vocative truncation of entire syllables is possible with some forms: (not possible for many speakers) Vocative truncation of entire syllables is widely attested outside of Indo-European as well. 17

[3.3] Non-vocative truncation
Truncation of names need not produce exclusively vocative forms, of course. In many languages, including Russian, truncation is used to make short, familiar forms of first names, and diminutive suffixes may or may not be added to the shortened forms, as in (9).
(9) Russian: English: Дмитрий Timothy Again, this is not an exclusively Indo-European phenomenon; truncation is used in Indonesian dialects, for example, to produce more familiar forms of first names as well (cf. Gil (2005)). All such truncation can be viewed as iconic for shortened distance between the speaker and the referent (who is the addressee in the case of truncated vocatives; see Yokoyama (1994)). When diminutive (or augmentative) suffixes are added to the shortened form in languages with productive diminutive (or augmentative) formation, such as Russian, the same form indicates both shortened distance toward the referent (conveyed by the truncation) and the speaker's particular attitude (e.g. affection) toward the referent (conveyed by the diminutive or augmentative suffix). 18 [16] On Sardinian see Floricic (2002), and on Romanian dialects see Maiden (2006, 52-53 English first (given) names are of course not as freely manipulated as Russian first names, and the use of shortened (familiar) or diminutive forms does not necessarily reflect the SPEAKER'S view of the discourse situation or attitude toward the referent, but often, rather, the NAME BEARER'S choice, so that a male with the name William may choose to always go by William or Will or Bill or Billy, etc., with very little variation "authorized", although this may evolve over time. The name bearer's choice can also be a factor in Russian when several short forms exist for a single given name.

.1] Czech
In Czech the vocative is formed by adding a vocalic ending to the stem, just like any other case, rather than truncating (or otherwise reducing) the nominative form, e.g. Věra > Věro!, Petr > Petře!. The vocative endings are the Czech reflexes of the historical Slavic vocative, and they are fully integrated into the nominal paradigm. Vowel-zero alternations are found in the vocative, just as with other cases, although since an ending is added, it is the zero alternation that is found in the vocative: In Czech, then, the vocative functions morphologically just like any other case (although case-status may not be admitted on theoretical grounds). 19 In Czech the vocative is used in all instances of direct address, for inanimates as well as animates; i.e. its usage is generally mandatory whenever a person or thing is being addressed. While it is certainly more common to address people and animals than inanimate objects, when inanimate objects are addressed, for whatever reason, the vocative is used (although a few noun classes have the vocative syncretic with the nominative), e.g. kniho! (<kniha) "book!" (Russian *knig! as a vocative is impossible), or hrnečku, vař! "little pot, boil!" (as in the Grimm tale).
Finally, the Czech vocative is used in all registers -formal and informal -and all discourse situations -serious and light -, although in informal speech it may not be marked on all components of the direct address form. This mainly concerns combinations with pán "Mr." (+ last name), where the last name may not receive vocative marking in informal speech, but vocative is nevertheless marked on pán > pane, e.g. pane Nováku! [vocative on both pán and Novák] / pane Novák! [vocative only on pán]. In contrast to the Russian truncated vocative, usage of exclusively nominative direct address forms is usually judged either impossible or marginal or rude in Czech. 20

[4.2] Polish
Polish presents a interesting contrast to both Russian and Czech. As in Czech the vocative in Polish is formed by adding special vocative endings (although, as mentioned above, truncated forms do exist dialectally). Unlike the situation in Czech, however, the vocative is not mandatory in Polish. According to Kottum (1983) and Anstatt (2003), in Polish the vocative is used for polite address in a formal [19] See Spencer & Otoguro (2005, 133ff) for a discussion of the theoretical status of the vocative in Czech and other languages. [20] My informants either ruled them out entirely or admitted them only as highly demeaning expressions of power of the speaker over the addressee, as though the addressee did not merit the speaker's recognition as an interlocutor, imaginable only in certain environments, such as the military. Such forms have, however, been attested elsewhere (cf. the opinion page in Naše řeč 1, ročník 26, 1942).
[218] lillian a. parrott or distant interlocutor relationship, where the nominative would be perceived as disrespectful or rude; this is almost the exact opposite of the usage of the truncated vocative in Russian. But the situation is in fact more complex: according to Anstatt, in closer or less distant interlocutor relationships, the vocative is also used, but only for addressing close friends or colleagues with whom the speaker is on friendly yet respectful terms. Kottum (1983), moreover, cites examples of the vocative being used in aggressive or insulting address in Polish, although the data are not entirely clear. In any case, vocative usage is marked in Polish, and it is attested on both ends of the address spectrum: for polite (distant) address, on the one hand, and for intimate (close) or rude address, on the other.

[4.3] Review
Russian thus differs from Polish and Czech in two main respects: (a) it has lost the reflexes of the historic Slavic vocative and instead uses truncation to make a special vocative form, and (b) the usage of this special vocative form is reserved for close interlocutor relationships -in its truncation the form is thus iconic for closeness, as noted by Yokoyama (1994). In Czech the vocative neutrally signals direct address, without any special added meanings. In Polish, however, standard vocatives are reserved for non-neutral address: more polite, distant, or respectful address at one extreme; and for insulting, friendly, or intimate address at the other extreme. 21 In short, the usage of the vocative in Polish is also a marked form as in Russian (and unlike Czech), but it can be used at both extremes of the address spectrum (close and distant). In Russian vocative truncation reflects the speaker's view of the interlocutor relationship as being close, in addition to overtly signaling direct address (like all vocative forms), but in Polish the vocative forms signal more generally some special awareness of the addressee on the part of the speaker -either that of respect, intimacy, or disrespect 22 -, again, in addition to signaling direct address. So far we have focussed on the peculiarities of truncated vocatives; now we will turn to direct address forms in general -not only specially marked vocative forms in Russian but also nominative-case direct address forms, and we will compare these to the usage of direct address forms in English. Direct address forms share a number of features across languages, but we will see further below that there are [21] Compare last-name direct address in English, which tends to be a sign of disrespect or intimacy (as in male camaraderie). [22] It is by no means rare that intimacy and disrespect are signaled by the same forms; cf. the usage of 2 nd person singular personal pronouns in languages that have a T/V distinction (Russian, French, German, etc.).

vocatives in contrast
[219] some important differences between English and Russian as concerns the usage these forms.
[5.1] Position • absolute (free): direct address forms can be used alone, in absolute position, just like interjections, as in (10a) and (11a), in which case they must be stressed.
• utterance-initial (left-adjoined or preposed): they can occur utteranceinitially, much like left-dislocation of arguments or various types of S-initial discourse markers (of which they are one; cf. Hey, Listen, OK, Right, Now, etc.), as in (10b) and (11b).
• medial: they can occur medially, i.e. interpolated at various points in an utterance, like other kinds of parentheticals, as in (10c) and (11c & d).

] Functions of direct addres forms: overview
In terms of function, direct address forms are used to attract, maintain, or focus an addressee's attention, and they can also serve to personalize an utterance in a variety of ways depending on the particular intonational realization. 23 The attention-getting function and the focussing or personalizing function are, however, different sides of the same coin: in all cases, direct address forms are addressee-oriented (like the imperative).

[5.3] Correlations between function and prosody
• absolute and utterance-initial direct address forms are stressed. Attention-getting direct address forms tend to occupy these positions. Although linguists may allude to "vocative intonation" or "vocative chant", a variety of quite different contours are possible, and these interact with stress, pitch, and vowel lengthening in subtle ways. 24 When the speaker already has the addressee's attention, an utterance-initial direct address form serves to maintain the contact and focus the attention on what follows, and absolute forms can assume a variety of other meanings (in addition to maintaining the addressee's attention) depending on the particular intonational realization (disapproval, begging, surprise, etc.).
• medial and utterance-final direct address forms tend to be deaccented or pronounced with low pitch, 25 like many kinds of parenthetical material. 26 Utterance-final direct address forms generally focus the attention on the preceding information, but they also personalize the utterance in a variety of ways, depending on the intonation and particular context. Medial direct address forms typically have a focussing function: they orient the addressee's attention to important information at the junction where they occur (i.e. preceding or following the direct address form), such as a preceding theme or a following rheme, or to the link between the preceding and following information. Medial direct address forms thus function like other parentheticals in that they can be interpolated at strategic points in [23] See Schegloff (1968); Zwicky (1974); Zaitseva (1992); Parrott (1995);Daniel' (2008); Daniel & Spencer (2009). [24] For example, calling contours on distal vocatives or direct-address forms (e.g. Maaaryyy!) are very different from utterance-final deaccented direct-address forms (e.g. I love you, Mary), and although some features are found cross-linguistically, languages of course differ in the ways that direct-address forms are realized prosodically. The intonational realization and prosodic integration of direct-address forms cannot, however, be addressed here; see Cruttenden (1986); Pierrehumbert & Hirschberg (1990); Ladd (1996); Gussenhoven (2004); Hock & Dutta (2010). [25] In some analyses utterance-final direct address forms are found to carry an independent (L*) pitch accent (e.g. Beckman & Pierrehumbert (1986); Pierrehumbert & Hirschberg (1990), whereas in others (e.g. Hock & Dutta (2010), and references therein) they are found to be deaccented. [26] On the prosody of parentheticals, see Dehé (2007), and references therein.

vocatives in contrast
[221] the host utterance, like linguistic flags marking important landmarks, in order to correctly orient and maintain the addressee's attention. 27 Direct address forms that appear to occur medially or even finally can, however, serve to (re)capture the addressee's attention, in which case they are stressed, and function like utterance-initial attention-getters, as in the English example below with the informal attention-getting particle hey: Although Russian and English direct address forms function quite similarly in many respects, they do differ in certain ways, especially as concerns medial position. Let us first look more closely at what non-initial direct address forms do.
[6.1] The function of direct address forms in non-initial positon Consider example (14); 28 Taisa Petrovna is Nikolaj's mother, and Nadja is Nikolaj's girlfriend. Ja 'I' cannot host mam(a) 'mom' in this particular context, and this has to do with what the speaker is trying to convey overall. The fact that he has just returned from the army is known to the addressee (the speaker's mother). What the speaker wishes to convey is that his mother should make an extra effort to be nice to his girlfriend since he just returned from the army, a difficult experience. But if the context is changed, as in (15), where what kind of person {I} is is at issue in the discourse, then ja/I becomes a possible host for the address form, both in Russian and in English. [27] On parentheticals see Dehé & Kavalova (2007). [28] This example is taken from Parrott (1995). [222] lillian a. parrott (15) Я, Мам(а), пришел из армии! Я человек бывалый. I, Mom, have come from the army! I am an experienced guy.
All the information in I, mom, have come from the army! is of course known to the addressee (the speaker's mother). What is at issue is what kind of person the speaker {I} is, and the special attention signalled by direct address form is thus sufficiently motivated in (15), whereas in (14b) it is not, and the medial vocative is not possible. Comparison of (14) and (15) suggests that medial direct address forms serve to punctuate or call attention to the preceding information in the utterance, which is thereby placed squarely in the center of the addressee's current concern, and there must be good reason to do so for that particular landing site to be acceptable, as shown by the unacceptability of (14b). But note that in (15) it is not simply the preceding information that is at issue, but the link between the preceding information -{I} -and the information that follows it -the trait {have come from the army} that defines the {I} here. The direct address form thus draws the addressee's attention to a junction in the utterance, and thereby highlights the surrounding information (not only the preceding information but also its link to the following information) and serves as a bridge between the two crucial pieces of information (usually corresponding roughly to theme and rheme). Note that parentheticals in general have been shown to have a similar punctuating or highlighting function (see Dehé & Kavalova (2007)).
With regard to utterance-final position, it may be the immediate constituent or the entire preceding utterance that is the object of special attention. Compare in this regard the following (made-up) examples: The speaker is unsure whether the lecture is tomorrow. Lunt's lecture is tomorrow, isn't it Mary? (↗) The speaker knows the lecture is tomorrow; she wishes to correct very subtly the addressee's (A) misguided assumption. A: This afternoon I'll be at Lunt's lecture, but in the evening I'll be free. a. B1: ???Lunt's lecture is tomorrow, isn't it Mary? (↗) b. B2: Lunt's lecture is tomorrow, isn't it? (↗) c. B3: Lunt's lecture is tomorrow, Mary.
In the meaning described in (16), no pause or shift in pitch direction occurs between the tag question and the direct address form, and the intonation rise continues from the tag through the direct address form. The direct address form can be attached to the tag in (16) because it is a truly informational (although biased) question. In (17a)-(17b), however, the tag is simply an indirect, polite way of correcting the addressee, and here, if the same contour is maintained (the in-vocatives in contrast [223] tonation rise continuing from the tag through the direct address form with no pause or shift), the tag cannot host the direct address form, as shown in (17a). The tag alone is fine, as shown in (17b), and if the tag is omitted, an utterancefinal direct address form is also acceptable, as in (17c); it is the combination of a non-informational tag question with the direct address form that is unacceptable. Thus, the function or information value of a final constituent can preclude postposed direct address marking.
[ Whereas medial position is possible for all the direct address forms in Russian, the equivalent English examples are all unacceptable or marginal, and yet the information structure is the same. So in English factors other than the information structure must also play an important role. Whereas Russian allows direct address forms to intervene between the verb and the direct object (19)-(21), English generally does not; such interpolation is particularly bad when the direct object is a pronoun, as in (20b). This is due in part to the more rigid word order of English, and to the tighter syntactic bonds between certain constituents. 30 But the interpolation of direct address forms causes particular problems, as the comparison with parentheticals given in the (b) versions below reveals: Another important factor in the interpolation of medial direct address forms concerns the prosodic structure and stress. Consider example (27), which is taken from Moro (2003): (27) a. There, Mary, is a solution to the theorem. (Moro 2003, 256) The only way this utterance can be acceptable is for there to be stressed, which means that we are dealing with locative there and not existential there, since existential there cannot receive stress. This is shown in (28) b. existential there (unstressed): *There, Mary, is a solution to the theorem. c. existential or locative there: Mary, there is a solution to the theorem. d. existential or locative there: There is a solution to the theorem, Mary.
As the data in (28) show, the host for a medial direct address form must be stressed (or at least stressable). Interestingly, there is not the same restriction on all syntactically external elements; for example, by the way (which is generally considered a parenthetical) is perfectly acceptable in this position, as shown in (29): (29) existential or locative there: There by the way is a solution to the theorem.
In the existential reading is would normally be stressed, and thus can serve as host to the parenthetical. But this option does not exist for the direct address form, which seems to require a stressed element to its left; this suggests that it is the preceding element that serves as the host to the direct address form. Note that in the discussion of examples (14)- (15) we saw that the informational status of the preceding constituent was crucial in determining possible landing sites for direct address forms, that only an informationally weighty constituent could serve as the syntactic host of an address form. In (28) as well we see that informationally poor and prosodically weak elements cannot normally host direct address forms.

[6.3] Syntactic and prosodic constraints on medial position in Russian
Although Russian, with its relatively free word order, distinct case marking, and high tolerance for scrambling, is very flexible when it comes to interpolating direct address forms, it does have some restrictions on the placement of medial vocative expressions, albeit to a lesser extent than in English, as shown in (30): (30) a. Бабушка, уважай вкусы других людей.
*Have some respect for other, Granny, people's tastes.
In (30b) the position after drugix is low in the syntax, and the insertion of babuška is judged impossible, whereas all other medial positions would be acceptable. Interestingly, however, if a more loaded direct address form were used, such as babulja, the utterance becomes acceptable, as shown in (31) [32] I wish to thank Elizaveta Khachatourian for this example. [226] lillian a. parrott As in the English examples, sufficiently loaded direct address forms can override syntactic constraints, as long as the information contained in the preceding constituent is in some way at stake in the context.
[7] c o n c l u s i o n In this article we have contrasted the usage of direct address forms in Russian with those in other languages. The Russian vocative is quite unusual in its formal and functional peculiarities, when compared to vocatives in other languages, such as Czech and Polish, which have vocative endings as opposed to truncated forms. Truncation as an iconic device to signal closeness or familiarity with the referent (or addressee in the case of vocatives and imperatives; cf. Yokoyama (1994)) is widespread, and in Russian it is particularly striking. Although the basic functions of direct address forms in Russian are the same as in other languages, Russian is shown to be particularly hospitable in its reception of medial vocatives as compared to English, which has considerably more syntactic and prosodic constraints on hosting direct address forms.
r e f e r e n c e s