1 Introduction: Analyses of authentic conversational interaction in phraseological research

Even though Burger (1998), Burger / Buhofer / Sialm (1982) and Fleischer (1982/1997)1 have delivered a comprehensive syntactic and semantic classification of phraseological expressions in German2, Elspass (1998) criticizes that

Arbeiten, die von einem geschlossenen Textkorpus als Materialbasis und empirischem Kontrollinstrument ausgehen, sind in der Phraseologieforschung immer noch die Seltenheit. (Elspass 1998: 26)3

However, this statement is no longer valid for the analysis of written German, where a great number of studies are based on corpora of literary (Christophe 1997, Palm Meister 1999), journalistic (Burger 1999, Wotjak 1992) or advertising texts (Mieder 1983, Coppens d’Eeckenbrugge 1999). On the other hand, it still has some truth in it as far as phraseological research on authentic oral verbal interaction is concerned: only few studies have been carried out in this area (Drew / Holt 1988, 1998 for English; Gülich / Krafft 1997, 1998 for French; Kallmeyer / Keim 1986, 1994; Keim 1997; Quasthoff 1983; Schmale 1999, 2001a, b, c, 2002 etc. for German). Consequently, little empirical insight into the conversational organization and/or the functions of fixed expressions (= FEs) has been gained so far. As a contribution towards filling this gap in empirical phraseological research a study based on 32 German talk shows (Arabella, Boulevard Bio, Doppelpass: DSF-Fußballstammtisch, Fliege, Ilona Christen, Birte Karalus, Hans Meiser, NDR-Talkshow, Jörg Pilawa, Bärbel Schäfer, Sonja, Andreas Türck, Vera am Mittag) was undertaken. Following James Schenkein’s (1978) credo of the "analytic mentality of the conversationalist", i.e. refraining from approaching data by using a pre-fabricated theoretical frame, in order to remain receptive for new, yet unnoticed organizational phenomena involving phraseological expressions, a phenomenon referred to as "the conversational treatment of fixed expressions" was discovered. In fact, a very thorough scrutiny of the taped talk shows revealed that a great number of phraseological expressions, especially those being idiomatic, and among these figurative (containing an image) and/or metaphorical (according to Burger 1998 the image has to correspond to a "concrete model"), are conversationally "treated" by the talk show host, but also by participants. "Treatment" of a FE implies that preceding, following or concomitant verbal and/or non verbal activities specifically relate in different ways to the phraseological expression in question. Based on detailed transcriptions4, involving segmental, suprasegmental and non verbal elements of (sequences of) utterances which comprise "treated" phraseological expressions, five main categories5 of conversational treatment of FEs were established:6

(I) Auto- and hetero-rephrasals of FEs, i.e. apart from necessary deictic modifications (in case of hetero-rephrasals), quasi-identical repetitions of the same FE (cf. Schmale 2001b)

(1) Talk show host Meiser asking Elke about her ex-boyfriend (HM/19-3/87)

HM sie sind erst ganz kurz single nech, ((…)) schade eigentlich sechs Jahre

 you split up not long ago didn’t you a pity really for six years he was

HM wars der Mann ihrer Träume, obwohl der immer auf ihrer Tasche lag,

 the man of your dreams even though you knew he was taking advantage of you

⇒ und dann haben sie (den) in die Wüste geschickt;

 but then you sent him packing

⇒ E dann hab ich ihn in die Wüste geschickt;

 yeah then I sent him packing

⇒ HM & ((lacht pp)) oder im Rhein versenkt-

 ((laughing a little)) or drowned him in the Rhine

⇒ E im Rhein versenkt ja;

 drowned him in the Rhine thats it

(II) Auto- and hetero-paraphrases of FEs, i.e. formally modified, but semantically more or less identical utterances relating to a (immediately) preceding turn (construction unit) (cf. Schmale 2007)

(II/1) Non-phraseological paraphrases of FEs

(2) A guest denies living in an illusory world (BS/19-3/3)

BS ((fragt ZU)) behauptest du einige Frauen machen ihren Job äh schlecht,

 ((host asking ZU)) do you mean that some women don’t do their jobs properly

⇒ ZU das kann ich nich beurteilen; ich sag nur jedem Tierchen sein Plaisierchen;

I can’t judge that all I’m saying is each to his own

⇒ jeder soll so leben wie er es für richtig hält.

 everybody should live the way he thinks is right

(II/2) Phraseological paraphrases of non-phraseological expressions

(3) Vera stating a well-known cliché about fat people (Vera/3-2/5)

V wenn man dick is is man langsam- is man faul- kommt man nich in die Pötte.

 when you’re fat you’re slow you’re lazy you never get anything done

(II/3) Phraseological paraphrases of another, preceding, phraseological construction

(4) Michael talking as a specialist on satanism (IC/3-6/5)

⇒ M: ein richtiger Satanist gibt sich normalerweise nicht zu erkennen;

satanists won’t disclose themselves

⇒ er agiert im Verborgenen- wie er schon gesagt hat;

 they operate in the shadows as he said

(III) Play on words with FEs (cf. Schmale 2005a)

(5) Host Meiser introduces his guest Mrs. Ortwein (HM/19-3/13)

HM es gibt tatsächlich so etwas wie:::- medizinische Wunder;

 they’re in fact things like… medical miracles

⇒ sie war 1984 total blind; (-) vor Liebe; inzwischen kann sie wieder sehn;

 in 1984 she was totally blind … with love but now she can see again

((Lachen im Publikum)) aus Berlin- Anneliese Ortwein herzlich willkommen.

((audience laughing)) from Berlin Anneliese Ortwein welcome to the show

(IV) Non-verbal activities relating to FEs (Schmale 2005b)

(6) Host Bio asks MR about her feelings when she found her father at last (Bio/2-2/5)

MR also ich sag ihnen ehrlich mir i::s- <<macht eine grosse Kreisbewegung>

 I can tell you quite frankly <<making a big circular movement with both hands>

SO ein Stein,> von meinem Herzen gefallen;

 it took such a load> off my mind

The 5^th category of "conversational treatment" encountered in the talk show corpus are

(V) Metalinguistic comments on and evaluations of FEs

(7) Host Hans Meiser (= HM) talking to his guests living on social security (HM/2-2/1)

HM wir zahlen ja alle Steuern; und von diesen Steuern leben sie;

we all have to pay taxes and you live on these taxes

⇒ und sie leben damit wie die Made im Speck sacht so mancher.

and you live like a bee in clover many people would say

The present paper concentrates entirely on this last type of "conversational treatment" which is in fact the only one of those mentioned providing definite proof that participants themselves consider a turn-construction unit as "fixed". Furthermore, by producing metalinguistic comments and evaluations of FEs, participants more or less explicitly display their attitude towards the FE employed, as well as their interpretation of type and communicative function(s) in different contexts.

The decision as to whether an expression has to be classified as phraseological or not was based on the criteria for the definition of FEs defined by Burger / Buhofer / Sialm (1982) and Burger (1998). An expression is thus regarded as fixed or phraseological in a wider sense when it is polylexical, i.e. when it consists of two or more lexical items (which may be auto-semantic or not)8 and when its form is stable, i.e. when it is generally used in this form as well as being structurally and psychologically "ossified".9 Fixed expressions in the narrow sense are those which are polylexical, stable and idiomatic, i.e. the phraseological meaning of the FE being different from its compositional meaning.10 While this third criterion is relatively simple to apply, stability is more difficult to determine as general use is not an objective matter. If lemmatization of an expression in a specialized dictionary for German (e.g. Duden 11, 1998) had been the only criterion for the classification of an expression as idiomatic, many of the polylexical constructions encountered in the corpus should have been neglected. However, in a great number of cases expressions in the corpus were classified as – at least potentially – phraseological in spite of their absence from Duden 11, mainly because of their idiomaticity, especially when they were metaphorical, but especially when they were accompanied by metalinguistic comments like wie man so schön sagt (as they say).

Following preliminary remarks on methodology (section 2), different types of metalinguistic treatment of FEs will be discussed: metalinguistic comments indicating that a polylexical expression is fixed (section 3.1.); the indication of the source of a quotation (3.2.); metalinguistic comments indicating producer-preferred meanings of FEs (3.3.); and finally metalinguistic evaluations of FEs (3.4.).

2 Methodological observations on the study of fixed expressions

One preliminary remark concerning the talk show corpus: even though one might argue that a highly institutionalized TV talk show does not qualify as everyday conversation11, which is normally characterized by the fact that all features of conversational organization can freely be negotiated by participants, TV talk shows – presenting non professional speakers using everyday language – were nevertheless chosen as it is highly unlikely that talk show producers or hosts induce participants to use or "treat" FEs in specific ways. If it is true that the story told by guests is known beforehand to talk show hosts and the overall structure of its telling is more or less strongly influenced by them, the latter also attach the utmost importance to spontaneity in order to create an impression of authenticity. It is thus more than unlikely that participants will have the exact wording of their conversational activities prescribed for them: neither the use of a specific phraseological expression nor any conversational treatment will be influenced by others. Furthermore, the study of face-to-face interaction has to be based on video material if non-verbal communication, which plays an important part in all areas of conversational interaction, is to be taken into account. Using professional television transmission of more or less spontaneous speech thus seems an acceptable solution.

Like Kallmeyer / Keim’s (1994), the presented study deals with the use of FEs in authentic conversational contexts.

Im Unterschied zu vielen Arbeiten zur Phraseologie, welche formelhafte Ausdrücke in Bezug auf ihre Integration in das Sprachsystem, insbesondere also ihre Lexikalisierung betrachten, untersuchen wir die Formelhaftigkeit primär unter dem Aspekt der Spachverwendung, d.h. als Formelhaftigkeit des Sprechens. (Kallmeyer / Keim 1994: 251)12

However, analyzing the use of phraseological expressions from a conversation analytic perspective has to take into account that the notion “fixed expression” is a pre-conversational concept13 defined by linguists following criteria of polylexicality, (structural and psychological) stability, frequency of use, and – in a narrow sense – idiomaticity. Like any other lexical item, FEs are thus used by participants as pre-fabricated turn-construction units in talk-in-interaction, but only exceptionally do they display orientation and relevance towards a FE as a particular type of turn-construction unit. This is the case when the producer of a FE himself – metalinguistically – (auto-)comments on an expression he uses or when another participant (hetero-)comments or (hetero-)evaluates a preceding expression as fixed or phraseological in one way or another. These metalinguistic comments or evaluations will be treated in the present paper as an ideal case of conversational treatment, ideal because the fact that an expression is fixed is genuinely being addressed by the producer himself or his co-participant.

But, in spite of the fact that the present study is based on a corpus of 32 authentic German talk shows and 229 transcribed sequences, one has to admit that in the great majority of cases of “conversational treatment” of phraseological expressions, “fixedness”, structure, meaning or use of FEs are not interactively negotiated by participants, which means that there are no traces of orientation towards the stable and/or idiomatic nature of an expression on the conversational "surface". Nevertheless, being aware of the existence of FEs from specialized dictionaries14 or from empirical studies15, analyses of corpora of authentic conversations can reveal conversational use and functions of FEs, which are not to be confused with communicative intentions attributed to the producer of a FE.16

Drew / Holt (1988) and (1998), whose studies are generally considered as part of conversation analysis strictu sensu17, do in fact start out from the idea that some (parts of) turn-constructional units are "figurative expressions", e.g. "between th’ devil ‘n deep blue sea"18 (Drew / Holt 1998: 516). On the grounds of empirically established stableness, they analyse their use in specific conversational environments and come to the following conclusion as to their basic conversational function:

[…] figurative expressions can be used to summarize and close down topics, and thereby to occasion a move to next topics. For this reason, we might regard the use of figurative expressions as an intersubjectively available practice or device for topic termination. (Drew / Holt 1998: 508)

The turn-construction units in question, however, considered by Drew / Holt (1998) as "figurative expressions", are at no stage addressed by the co-participants themselves in any specific way as FEs.

Kallmeyer / Keim (1994), studying stereotyped in-group language in the Filsbach area of Mannheim, believe that "formulaic" speech can be successfully distinguished from "normal" speech by studying prosodic and textual phenomena:

So indizieren vielfach prosodische Merkmale einen Kontrast der formelhaften Äußerung zum Kontext und markieren durch Auffälligkeiten der Sprechweise auch die Reproduziertheit. (Kallmeyer / Keim 1994: 257)19

This may be the case in Kallmeyer / Keim’s (1994) corpus, where group-specific FEs are being used in sequences frequently marked by a high degree of emotional involvement, but this observation is not confirmed by the use of FEs in the corpus the present study is based on. In any case, prosodic characteristics like strong word accent, loudness or higher pitch are not distinctive features for the recognition of a FE, as they might well be used for isolated lexical items, for instance to assure the comprehension of a difficult technical term whose production is preceded by a micro-pause, slow and particularly careful pronunciation and louder speaking. For speakers in Kallmeyer / Keim’s (1994) corpus, the FEs used may have a particular in-group status, as they contribute to the definition of the group’s identity in a wider sense. In the talk show corpus, however, FEs are used like "normal" lexical items, certainly for specific reasons, but they are not prosodically distinguished from their environment.

The analyses undertaken in the following sections will thus be limited to different types of metalinguistic comment and evaluations of FEs mainly on the segmental level.

3 Metalinguistic treatment of fixed expressions

Metacommunicative, metadiscursive or metalinguistic20 verbal and non verbal activities are examined by linguists as a general phenomenon of language use (cf. Authier-Revuz 1993; Gülich 1994; Gülich / Kotschi 1995; Morel 1985; Techtmeier 2001), and as a specific feature related to commenting and evaluating phraseological expressions (cf. Bastian / Hammer 1998, 1999, 2000a, 2000b, 2002; Dobrovol’skij / Lubimova 1993; Schmale 2002; Wotjak 1992).

Gülich / Kotschi (1995: 51-57) describe metalinguistic evaluations or comments as a third type of discourse production apart from verbalization and treatment activities, covering "a large group of expressions (of a more or less explicit) metalinguistic or metadiscursive kind" (51) which are employed by speakers "to evaluate or to comment on expressions (or sequences of expressions) which are part of the ongoing discourse" (51).

Burger et al. (1982) point out that phraseological expressions in particular, proverbs and "other types of phraseological expressions" are almost systematically accompanied by comments of the as-they-say-type.

Sprichwörter (kommen) heute kaum mehr ohne metasprachliche Kommentierungen von der Art wie man so schön sagt oder es gibt da ein Sprichwort (vor). Bei anderen phraseologischen Typen kann durch vergleichbare Signale auch das ‘Redensartliche’, das ‘Gebräuchliche’ der Wendung, auch die Tatsache, daß man eine derartige Situation üblicherweise so und so charakterisiert, angezeigt werden. Oder es wird darauf hingewiesen, daß man sich des metaphorischen, jedenfalls nicht-wörtlichen Charakters der Wortverbindung bewußt ist, daß man sie allenfalls im phraseologischen und/oder wörtlichen Sinn meint u.ä. (Burger / Buhofer / Sialm, 1982: 89)21

In the talk show corpus analyzed, three main categories of metalinguistic treatment of FEs were discovered which will be discussed and analyzed as to form and functions in the following sections.

3.1 Metalinguistic comments indicating that a polylexical expression is fixed

3.1.1 Metalinguistic terms

In the corpus terms like (Wahl-)Spruch (motto), (Lebens-)Motto, Phrase (hackneyed phrase), Floskel (set phrase) etc. are used to describe certain expressions (which can precede or follow the term) as fixed and/or generally known. There is only one instance of metalinguistic comment which contains a term used in phraseology, i.e. Sprichwort (proverb) or, to be precise, Spielwort, as the non-native speaker of German puts it, but context clearly supports the view that what was meant was Sprichwort: in alle Sprachen dieser Welt gibt es ein Spielwort auch in Deutschland.

(8) Nadia (= N) has a very strong Eastern European accent (Vera/3-2/6)

⇒ N in alle Sprachen dieser Welt es gibt ein Spielwort auch in Deutschland- (.)

 in all languages of this world there is a proverb also in Germany

⇒ Liebe ist blind; Liebe kommt von der Seele; is egal wer ist das;

 love is blind love comes from the soul it doesn’t matter who it is

wie ist das Aussehn; wenn er liebt seine Frau oder sie liebt ihren Mann,

 what she looks like if he loves his wife and sheloves her husband

sie brauchen nicht zu suchen ob er ist dick oder dünn, das spielt keine Rolle.

 they don’t have to look if he is fat or slim it doesn’t matter

It is of no importance for the analysis that both the metalinguistic term (Spielwort for Sprichwort/proverb) and the proverb that follows are erroneous (the correct German version Liebe macht blind suggests a process, unlike English Love is blind)22. The relevant fact is that the speaker explicitly characterizes an expression as proverbial, and not just in German but in any language (a statement that she does not have to prove).

Further instances of metalinguistic comment contain words used in everyday language such as Spruch, implying that the expression referred to has a fixed shape, that it is regularly used by at least one person, that it has a concise and expressive form (and may even be considered well-constructed).

(9) Lesbian Andrea (= A) talking to "hetero" guests (IC/2-2/13)

⇒ A aber ich denke einfach mal- kennt ihr nich den Spruch leben und

but I just think don’t you know the expression live

⇒ leben lassen? (-) lasst sie doch; lasst sie doch einfach.

and let live just let them just let them

The metalinguistic comment kennt ihr nich den Spruch (don’t you know the expression) implies on the one hand that the expression leben und leben lassen (live and let live) should be known to listeners; on the other hand, because of the term Spruch, it implies that the commonplace, marked by triple alliteration, has a fixed form and is commonly used in this way.

The function of the metalinguistic activity, indicating the phraseological nature of the expression qualified, is obvious in the two sequences quoted: speakers N and A explicitly refer to a generally known FE, which is quoted as a sort of reference, summing up experience that cannot be contested and which thus strengthens their arguments.

3.1.2 Metalinguistic comments implying common use

The most frequent type of metalinguistic comment is represented by the combination of a non-specified generic agent, e.g. man (one) or a partial quantifier like mancher (many/some people), and a verb of saying, indicating that the expression referred to is commonly used.23 Here are two instances of this type:

(10) Young delinquent Thomas (= T) has told Fliege (= F) that he is ashamed of what he did to his victims (Fliege/2-6/3)

⇒ F dann sacht ma:n- ähm (n) schlechtes Gewissen is nich n gutes Ruhekissen

and then people say erm (a) bad conscience stops you from sleeping well

⇒ also- (.) geht die:se Geschichte mit einem in die Nacht rein, (--)

I mean this story does it follow you into your sleep

also träumst du davon,

I mean do you dream about it

(11) Host HM talking to his guests living on social security (HM/2-2/1)

HM wir zahlen ja alle Steuern; und von diesen Steuern leben sie;

well we all pay taxes and you live on these taxes

⇒ und sie leben damit wie die Made im Speck sacht so mancher.

and you live like a bee in clover many people would say

Both dann sacht man (then people say) in (10), preceding the highly modified proverb24 Ein gutes Gewissen ist ein sanftes Ruhekissen (those who have a clear conscience sleep well), and sacht so mancher (many say) in (11), following the verbal idiom wie die Made im Speck leben (to be in clover/ to live in luxury), refer to and comment a FE. Even if both FEs are well known, hosts Fliege and Meiser point out, via a preceding or following metalinguistic comment, that the FEs are employed by members of the speech community, which implies at the same time that they are more or less stable. Why do F and HM engage in this activity? Certainly not for the same reasons as in sequences (8) and (9) in 3.1.1., where the producers of the metalinguistic terms were trying to convince their audience of the argument developed. In (10) and (11) the function of dann sacht man/sacht so mancher is in some way self-protection, especially in the second case.25 The contents of Meiser’s turn – you live extremely well without doing any work just by relying on the taxpayer – could be interpreted as strongly face-threatening by the addressees, and Fliege’s turn, implying that his guest must have a guilty conscience, to a lesser extent, too. Referring to the common use of the FEs in question thus transfers the responsibility for their contents from the speaker to the speech community, enabling the individual producer to counter a possible complaint by: I did not say you were living in luxury, but many people think so. At the same time the producer of the FE demonstrates his communicative competence by meta-communicating about his own turn and by displaying his familiarity with the ossified character of the expression.

Further examples of this type are was sich liebt das neckt sich oder wie war das (how do you say – teasing is a sign of love) or die Geschichte mit dem berühmten Fass (the well-known story of the last straw)26.

3.2 Indication of the source of a quotation

The producer of a FE can indicate its exact source27. In our first extract of this type this is the Bible.

(12) Host Bio talking to comedy star Hella von Sinnen (= HS) (Bio/16-3/1)

HS wenn du geliebt wirst is auch alles in Ordnung;

well if you are loved everything is fine

und ich liebe mich natürlich sehr; aber ich- ((Publikum lacht))

and of course I love myself very much but I ((audience laughing))

⇒ Bio das is ja sehr biblisch; (-) in der Bibel steht liebe deinen Nächsten

that’s very biblical it says in the Bible thou shalt love thy neighbour

⇒ wie dich selbst; offensichtlich is der Maßstab-

as thyself apparently it is very important

die Liebe die man zu sich selbst (hat) was viele vergessen.

the love you feel for yourself this is what many people seem to forget

HS das find ich sehr klug; (.) da hab ich noch nich drüber nachgedacht.

I think that is very clever I have never really thought about that

⇒ Bio ja das is so; das steht da so.

yes that’s so that it how the Bible puts it

In commenting on HS’s ich liebe mich natürlich sehr (of course I love myself very much) by das is ja sehr biblisch (that’s very biblical), Bio explicitly gives the quotation from the Bible which HS’s turn makes him think of, i.e. liebe deinen Nächsten wie dich selbst (thou shalt love thy neighbour as thyself); he introduces it via the metalinguistic expression in der Bibel steht (it says in the Bible), thus indicating the source28.

The source of a FE can also be a well-known personality, as in the following two sequences from DFS football talk.

(13) Football manager Rainer Bonhof (= BO) at a press conference (Dopa/7-3/7)

⇒ BO man müsste glaub=ich ä::hm den Satz von Bernhard Langer (--) benutzen,

one would have to I think use Bernhard Langer’s observation

⇒ dass Leistungssport sich zwischen den Ohren abspielt;

that nowadays top level sport is a matter of what goes on between the two ears

RB expressly attributes the metaphorical FE29 – Leistungssport spielt sich zwischen den Ohren ab (top-level sport is a matter of what goes on "between the two ears", i.e. is a question of psychology) – to German golfer Bernhard Langer30 via the metalinguistic comment den Satz von Bernhard Langer (BL’s sentence/observation).

And in the following extract (14), RB does the same for Werder Bremen’s Manager Willi Lemke who likes to use (wie er immer sagt) the expressions Pickel kriegen und eine schreckliche Version sein.

(14) Host RB talking about Bayern Munich’s dominant role (Dopa/7-2/10)

RB der FC Bayern soll unser Thema sein noch zum Abschluss, und zwar-

to finish with we want to talk about Bayern Munich I mean

wo wir schon mal den Präsidenten hier haben,

as we have the president with us

müssen wir natürlich auch über den FC Bayern reden,

we have to talk about Bayern Munich as well

und es is ja die Übermacht der Liga sagen zumindest äh die Konkurrenten;

and it is their supremacy in the league at least that’s what their rivals say

⇒ unschlagbare Bayern, da kriegt Willi Lemke Pickel,

invincible Bavarians it brings Willi Lemke out in a rash

⇒ beziehungsweise für ihn ist das eine ganz schreckliche Ver’

or rather for him this is a

⇒ Vision wie er immer sagt.

horror vision as he always says

Whereas the expressions in (13) and (14) have actually been employed by two widely known protagonists from professional sports, the following one in extract (15) – wir müssen alles analysiern –, is hypothetically attributed to German football international Marco Rehmer. However, not because he is known to have used it before, but because another well-known footballer, Andreas Möller, has done so. It thus seems that wir müssen alles analysiern, as opposed to Leistungssport spielt sich zwischen den Ohren ab (13) and Pickel kriegen (14), has more of a genuinely fixed expression which is not just limited to one person, as it is more or less frequently used at least by a group of people, in this case by professional footballers.

(15) Host Rudi Brückner (= RB) asking Franz Beckenbauer (= FB) why defender Marko Rehmer made so many disastrous mistakes (Dopa/7-2/2)

RB der Mann weiß doch wie Fußball gespielt wird warum steht der da,

the man knows how to play football so why does he just stand there

ich frags noch mal;

let me ask the question again

FB am Besten is sie richten die Frage direkt an ihn.

the best thing would be to ask HIM the question

⇒ RB ja dann sagt er wir müssen alles analysiern;

in that case he’ll answer we have to analyze everything

⇒ das hat Andy Möller letztes Mal auch gesagt.

 that’s what Möller said last time too

The question of classifying the expressions in (13), (14) and (15) as fixed deserves some attention. Quite obviously all of them are considered as typical of the people they are attributed to. The fact that someone regularly uses a specific expression (cf. (13) or (14)) or will probably employ a certain type of phrase (cf. (15)), indicates stability of form and some currency. However, do we have to consider expressions of this type as fixed when we know that they will most likely never find their way into a dictionary? It seems nevertheless that these short-lived and group-specific expressions have to be regarded as fixed, as continuous existence over a long period and wide-range usage are not amongst the obligatory defining criteria of phraseological expressions. An expression can thus be fixed even if it is used by few people and exists only for a limited period. The concept of stability would probably greatly benefit from being linked to "situational fixedness", similar to routine formulae or pragmatic idioms, which are totally context-dependent. Having for a long time worked without coherent corpora (cf. Elspaß 1998: 26) in general and without looking into the functions of FEs in authentic dialogue in particular, phraseological research has not been able so far to take into account "situational fixedness"; this does, however, deserve the attention of the analyst, considering the indexicality of all speech.

The communicative functions of the metalinguistic comments in (12) and (13) are not identical: in example (12), Bio has to or at least chooses to quote the exact source of the saying in order to prove that his evaluation of HS’s sentence as "biblical" is correct. At the same time he demonstrates his knowledge of the Bible, his status as host (who has the right to comment on guests’ contributions) and his own communicative competence. The reason for the indication of the source in (13) is a different one. To avoid being accused of plagiarism speakers obviously do not want to use a highly original expression without indicating that it has been created recently by someone else: credit where credit is due! And as in the case of functions of metalinguistic comments implying common usage of an expression (cf. 3.1.2.), this technique enables the speaker to keep his distance from the expression quoted: it is Bernhard Langer who says so, not I. Of course, he also demonstrates his communicative skill by using a highly figurative expression, even if it is not his own invention, to concisely express an extremely complex state of affairs.

3.3 Metalinguistic comments indicating producer-preferred interpretations of FEs

The most common type of metalinguistic comment concerns techniques for indicating or demonstrating speaker-preferred meaning of a FE and thus guiding or even influencing hearer-interpretation.31 Wotjak (1992: 126ff.) points out that metalinguistic units like im wahrsten Sinne des Wortes, buchstäblich, wie man so schön sagt, sprichwörtlich, etc., preceding or following "phraseolexemes", function in this way:

Diese Wirkung kann u.a. darin bestehen, dass der Rezipient darauf hingewiesen wird, dass sich der Sender nur bzw. auch auf die Ebene der freien Wortverbindung (…) bezieht oder dass er im Text nur die phraseologische Bedeutung (…) und nicht auch bzw. an ihrer Stelle die (freie Wortverbindung) der homonymen freien Wortverbindung, wie sie im Kontext möglich schiene, akzeptiert wissen möchte. (Wotjak 1992: 126)32

Naturally any utterance contains clues as to what kind of hearer-interpretation the speaker is trying to achieve, e.g. a specific syntactic structure or rising intonation to produce a turn conventionally interpreted as a question by the hearer (do you want to go to the cinema?), but in the metalinguistic comments dealt with in this chapter the producer of an FE makes his preferred33 interpretation more or less explicit. The speaker knows that the hearer can in no way be forced to respect his intended meaning, as the producer of an utterance is by no means master of its interpretation. In fact, the interpretation of an utterance can occasionally turn out to be totally opposed to the speaker’s initial intention! Four types of comment indicating speaker-preferred interpretation of FEs will be distinguished.

3.3.1 The in-quotation-marks-type

The metalinguistic comment in Anführungsstrichen/-zeichen (in quotation marks)34, in itself a routine formula, which can also be expressed non-verbally by one or both hands being held up to draw inverted commas in the air by two slight upward-downward movements of the index and middle finger, may refer to phraseological as well as to non-phraseological expressions. In the examples presented this – verbal – comment-type precedes the FE, but it could also follow it.

(16) Host RB asking a question about footballer Lothar Matthäus (Dopa/7-2/7)

RB also is doch der FC Bayern schon n Schritt weiter als die Nationalmannschaft,

but this means that Bayern Munich has gone a step further than the national team

und das ( ) geht wieder n Schritt zurück und holt einen Lothar Matthäus,

who are taking a step backwards by appointing Lothar Matthäus

⇒ der beim FC Bayern so langsa:m- (--) <<Handbewegung mit beiden

who is progressively <<movement with both

⇒ erhobenen Händen> in Anführungsstrichen,> aufs Altenteil (.) gesetzt wird.

hands raised> in inverted commas> regarded as being due for retirement

The verbal idiom jemanden aufs Altenteil setzen (make someone retire)35 is preceded by verbal and nonverbal quotation marks. However, the nonverbal quotation marks do not frame the expression itself; they rather accompany the expression in Anführungsstrichen (in inverted commas). There is not one single instance in the corpus where nonverbal inverted commas are opened with the left hand before the expression in question and closed with the right hand at the end of it.

In two further sequences, (nonverbal) inverted commas are used in the same way.

(17) Host Sonja (= S) summing up Anja and Kai’s problems (Sonja/29-5/6)

S also zu euch beiden kann ich jetz einfach nur sagen Eifersucht kann,

well the only thing I can say to you (two) is jealousy can

⇒ natürlich ich sag jetz mal <<Geste mit erhobener linker und rechter Hand>

of course be let me say <<gesture with raised left and right hand>

⇒ in Anführungsstrichen, > ein Zeichen von Liebe sein,

in inverted commas> a sign of love

aber es soll natürlich nicht einengen.

but of course it shouldn’t restrict your freedom

(18) Michelle (= M), talking to host Birte Karalus (= BK), was herself very young when she had to take charge of her sister’s education after their mother’s death (BK/2-2/3)

⇒ BK Michelle war das für dich ganz normal diese Mutterrolle einfach zu übernehmen,

and Michelle was it perfectly normal for you to simply assume this mother’s role

⇒ ((one cannot see BK)) in Anführungszeichen dicken Anführungszeichen,

in inverted commas thick inverted commas

⇒ einfach die Mutterrolle zu übernehmen,

just to take over your mother’s role

The collocations ein Zeichen von … (a sign of…) and die …rolle übernehmen (assume the role of …) are preceded by the same routine formula, i.e. Anführungszeichen (inverted commas) as in (16); in (18) BK even repeats it and reinforces it with the adjective dick (big fat).

The function of this metalinguistic phraseological comment is more or less the same in all three cases, i.e. the FE that follows is not totally appropriate to describe the state of affairs intended by its producer or the implications or connotations are too strong and/or not appropriate for the situation. Even if footballer Lothar Matthäus is indeed replaced by a younger player in his club, he is still only thirty-eight years old at the time of retirement from professional football, whereas the idiomatic expression aufs Altenteil setzen is normally reserved for people over sixty or even seventy retiring from professional life.

In (17), host Sonja obviously thinks that jealousy (Eifersucht) is not a sign of (true) love (ein Zeichen von Liebe) and thus puts the FE into inverted commas in order to demonstrate this to her audience. It is probable the highly conflictual relationship between guests Anja and Kai which makes host Sonja indicate that the FE ein Zeichen von Liebe is not to be interpreted literally.

The same observation applies to (18): Michelle was only sixteen when she took over the role of mother to her baby sister. However, the compound noun Mutter/rolle precisely expresses that the person referred to assumes the mother’s role without being the real mother. The expression is thus perfectly adequate, unless BK wants to qualify the lexeme mother. If not, it is not impossible that host Birte Karalus produces an illogical comment, typical of talk shows, thinking nevertheless that she is presenting herself as a competent speaker of German.

To sum up the basic function of (nonverbal) inverted commas as metalinguistic comments preceding a phraseological expression one could say: the phraseological sense of the FE I am using is not totally appropriate to the situation because the FE concerned has connotations and implications that do not apply to the state of affairs described.

3.3.2 The if-I-can-put-like-that-type

In example (17) we already encountered the routine formula ich sag jetz mal (if I can put it like that)36 which is reinforced by nonverbal and verbal inverted commas. It is a fashionable expression, frequently used, not as much as in French, however, where even news speakers on television have been heard using on va dire.

(19) S reintroducing the last subject discussed after commercials (Sonja/3-2/4)

S jetz is=es ja so man kann überhaupt nich glauben dass man in einer Ehe- (--)

now the thing is it is very difficult to believe that within a marriage

fünfzehn Jahre lang es nich mitbekommen kann dass jemand- (--)

one has not noticed for fifteen years that somebody

⇒ ja ich sag jetz mal ne Leiche im Keller hat;

if I can put it like that has got a skeleton in the cupboard

einfach (.) Drogen nimmt- es kann ja auch immer irgendwas andres sein;

it’s simply that he takes drugs or it can be all sorts of other things

dass man so was überhaupt nich WEISS von seinem Partner.

that you have absolutely NO IDEA about this side of your partner’s life

Extract (20) contains another instance of this type.

(20) The host has asked D how he reacted to his wife’s putting on weight (JP/3-2/1)

⇒ D ick hab auch keene Freudensprünge gemacht

I didn’t really jump for joy

⇒ (weil mir das) nich gefallen hat sagn wa mal so;

(because) I wasn’t very pleased about it let’s put it that way

The function of this if I can put it like that comment is very close to the function of inverted commas, but unlike the metalinguistic comments in 3.3.1., this metalinguistic expression not only indicates that the FE referred to should not be taken literally, it also implies that the expression used is probably not totally adequate, precise or as well chosen as it could be. While in French it seems to be a "figure méta-énonciative du ‘bien dire’" (Authier-Revuz, 1993), it would be more appropriate in German to describe it as a metalinguistic formula of approximative or imprecise, probably of sloppy, speech. A comment heard on television – sagen wir mal so salopp (let us say loosely) – confirms this interpretation: the producer states that he has not taken as much care in constructing his turn as he should have.

But why would a speaker make such a comment, endangering his own face? Most likely because, even if he indeed does so, the advantages are stronger than the disadvantages. On the one hand, the speaker’s metalinguistic comment can be interpreted as an activity aimed at creating reciprocity, i.e. S is aware of the fact that he is neglecting the conversational maxim of manner ("Be perspicuous!") and explicitly admits this to H: I know I am not expressing myself as precisely as I should, but I cannot do any better at the moment. On the other hand, the producer presents himself as a competent speaker, able to be critical of his own speech.

3.3.3 The so-called-type

Two lexemes – sozusagen and sogenannt – also serve the function of metalinguistic comment.37

(21) Host Vera (= V) talking to football supporter Dirk (= D) (Vera/29-5/1)

V wie weit würdet ihr gehn, (.) um euer Fandasein in (.) irgendeiner Form- (--)

as fans how far would you go to show your support

durchzusetzen; (--) würdet ihr euch prügeln- (.) Diskussionen anfangen-

would you fight start arguments

wie weit würdet ihr gehen Detlef und Thomas;

how far would you go Detlef and Thomas

D Diskussionen auf jeden Fall;

arguments most definitely

⇒ D aber die sogenannte dritte Halbzeit also Gewalt bringt absolut nischt.

 but the so-called third half-time I mean violence doesn’t makeany sense at all

The adjective sogenannt (so-called), which always precedes the expression referred to, indicates that the FE die dritte Halbzeit (the third half-time of a football match) does not have to be taken literally and, at the same time, that it is a stable expression, well known within the relevant part of the speech community. However, if it is a FE, characterized by frequent use in a stable form, then why does S have to state this? Probably because he is not sure that everybody understands the expression. This interpretation is confirmed by D’s paraphrase – also Gewalt (I mean violence) – which explains the meaning of the idiomatic expression.

The next extract provides an example of sozusagen.

(22) Host A wants to know why L is so much in love with her boyfriend (Ara/30-6/3)

⇒ L es gibt niemanden der irgendwie::- (--) ihm das Wasser reichen kann sozusagen;

there is absolutely nobody who somehow who can hold a candle to him so to speak

A was ist denn so toll an ihm,

what is it that is so fantastic about him

L das kann man schwierig irgendwie erklären; weil er is einfach mein Traumtyp;

that’s very difficult to explain because he’s just the man of my dreams

If sozusagen has basically the same function as sogenannt, i.e. the indication of non-literality, in this instance it probably expresses two more aspects: firstly, Lil does not use the FE jmdm nicht das Wasser reichen können (someone is not fit to hold a candle to someone else) in its real meaning which, according to Duden 11 (1998: 783) is "not have the same capacities as someone else". As a matter of fact, it is not Lil’s boyfriend’s capacities in general that are aimed at but his ability to attract her love in particular. Secondly, Lil possibly finds the expression inappropriate for somebody very close to her, as one belonging to her "negative face", to her personal territory, in Goffman’s terms.

Unlike the adjective sogenannt the adverb sozusagen (so to speak) can precede or follow the expression it qualifies. Another difference is that whereas sogenannt always comments on a FE or a well-known monolexical expression, sozusagen can also be used to mark a non-conventional unique or personal use of a lexical item.

3.3.4 The "literally"-type

As opposed to the markers of non-literality (cf. 3.3.1., 3.3.2. and 3.3.3.), there are those like im wahrsten Sinne des Wortes (literally)38 or auf gut Deutsch (in plain words) that emphasize literality (cf. Gréciano, 1983: 252).

For Burger / Buhofer / Sialm (1982) this type of metalinguistic comment, which itself takes the form of a FE, indicates

dass die kommentierte phraseologische Wortverbindung besonders ‘treffend’ und ‘sinnvoll’ verwendet, weil klar motiviert ist. (Burger / Buhofer / Sialm 1982: 89)39

In the following example talk show host Alfred Biolek classifies his guest’s FE aus heiterem Himmel (out of the blue) in this way.

(23) Host Alfred Biolek (= Bio) talking to guest Bärbel Schäfer (= BS) (Bio/2-2/1)

Bio vor vier Monaten kam ihr Partner bei einem Verkehrsunfall ums Leben;

four months ago your partner was killed in a car accident

BS mhm,

yeah

⇒ Bio ja da:s- (--) [letztendlich40] aus heiterem Himmel;

⇒ BS <<leise> aus heiterem [Himmel.]>

<<low> out of the blue >

⇒ Bio im wahrsten Sinne des Wortes.

in the true sense of the term

First Bio rephrases BS’s FE aus heiterem Himmel (out of the blue), then he produces his metalinguistic comment im wahrsten Sinne des Wortes (in the true sense of the word). It seems that the function of the metalinguistic comment is to underline the FE, to draw attention to it rather than indicating its literal nature; and of course to demonstrate that the FE is well adapted to the situation.

However, im wahrsten Sinne des Wortes can have a literal meaning, as in the following sequence.

(24) Host Rudi Brückner’s (= RB) comment on Rostock’s manager (Dopa/14-3/6)

⇒ RB ((direkt nach einer eingespielten Kurzreportage)) ha:t Ewald Lienen

((after a short report)) doesn’t have Ewald Lienen

⇒ im wahrsten Sinne des Wortes nicht mehr auf dem Zettel,

on his list any more in the true sense of the word

und äh wir reden gleich über (.) diese (.) Loyalitätsfrage nach der Werbung;

and erm we’ll talk about this question of loyalty in a minute after the commercials

Manager Lienen is known for continuously taking notes on small pieces of paper during a match. The metalinguistic comment on the expression jemanden (nicht) mehr auf dem Zettel haben (not have someone on the list any more), variation of jemanden (nicht mehr) auf der Rechnung haben, seems to refer to Lienen’s habit as well as to his not being on Hansa Rostock’s payroll any more, having been sacked after a series of defeats.

auf gut Deutsch (in plain words)41 is used with FEs, but also non phraseological expressions or lexical items, which may be considered as crude or vulgar, but which perfectly express what its producer wishes to state.

(25) Host Alfred Biolek (= Bio) asking guest Holger Müller (=HM), imprisoned for being suspected of paedophilia, about his relationship with his fellow inmates (Bio/30-6/2)

Bio man is in der Hierarchie der Gefangenen ganz unten;

you are at the very bottom of the prison hierarchy

⇒ HM ja auf gut Deutsch man- (.) man is der letzte Dreck.

yes in plain words you they treat you like dirt

Guest Holger Müller thus introduces his paraphrase man ist der letzte Dreck (they treat you like dirt) of Bio’s man ist in der Hierarchie ganz unten (you are at the very bottom of the hierarchy) by auf gut Deutsch in order to show that he is aware of using an expression which is probably totally appropriate to the situation, but may not correspond to the hearer’s expectations because it may be felt to be rather crude.

3.4 Metalinguistic evaluations of FEs

Apart from the stereotyped forms of metalinguistic comment described and analyzed in the previous chapter, which indicate producer-preferred interpretation, both the speaker himself and the hearer can produce metalinguistic evaluations of speakers’ FEs.

(26) Guest Kurt Bittner’s (= KB) about his ex-wife’s greediness and host Hans Meiser’s (= HM) comment on it (HM/19-3/5)

⇒ KB wiewiel Zucker soll da noch in das berühmte Hinterchen gesteckt werden?

how much more sugar does one have to blow up her bottom so to speak

⇒ HM ja nun mal vorsichtig hier ja, (is ja) ne jugendfreie Sendung hier.

hold on careful this is a programme watched by children as well

KB: Nase voll.

I’m fed up with it

KM’s moderate version42 of jmdm Zucker in den Arsch blasen43 (literally: blow sugar up somebody’s arse), i.e. do everything to please someone, is criticized by Meiser who asks his guest to mind his language because the programme might be watched by children. KB’s FE is thus evaluated negatively as not appropriate for the show.44 But it has to be pointed out that Meiser does not reject the expression used by his guest because of its fixedness but rather because of its vulgar nature; a comment of this type could thus also concern non-phraseological expressions, e.g. four letter words.

On the contrary, the negative or rather humorous evaluations observed in a talk show on football, the "Doppelpass: DSF-Fußballstammtisch", directly refer to the stableness of FEs. In this programme, broadcast on DSF every Sunday morning between 11 a.m. and 1 p.m., participants who produce Phrasen (i.e. expressions regarded by the host as hackneyed or worn-out) have to pay three euros45 into a Phrasenschwein (piggy bank) when the host of the show so decides, without there being a precise definition of what exactly a Phrase (hackneyed phrase) is.

(27) Stuttgart’s president Meyer-Vorfelder (= MV) talking about replacing the team-manager and the effects of this; host Rudi Brückner’s (= RB) reaction (Dopa/7-3/6)

MV dann geht der Trainer irgendwo anders hin und da klappt es wieder; (--)

and then the manager moves to another club where he is successful again

⇒ und dann schweigt man oder sagt ja ja is gut neue Besen kehren gut;

 and nobody says anything or they might sayalright new brooms sweep clean

⇒ RB ja- (--) da muss ich Herrn Finanzminister bitten;

well for this (expression) I’m afraid I’ll have to ask you (to pay)

MV ich muss noch mal wiederholen auch wenn ich jetz da was zahlen muss, (--)

but I’ve got to say it again even if I have to pay

⇒ das sind ja auch immer äh äh (.) wirklich ausgelutschte Floskeln

it’s always erm erm really worn-out phrases

⇒ die da verwendet werden mit den Trainern, das (.) hängt eben (.) davon ab

which are used about football coaches it depends you see

kann ich wirklich sagen ob ein Trainer in der konkreten Si’

and I can really confirm this on whether in a given

Situation zur Mannschaft (.) passt (.) das ist das- (.) Entscheidende.

 situation a coach is right for the team that’s what’sreally important

MV’s Neue Besen kehren gut46 (New brooms sweep clean) is indirectly evaluated by host Rudi Brückner as a Phrase and consequently MV has to pay. MV himself then produces an evaluation of his FE, i.e. ausgelutschte Floskeln (worn-out phrases). The criterion for classifying an expression as a Phrase thus seems to be the frequency of use in a specific situation, i.e. statements on football are penalized when considered hackneyed, worn-out or over-used.

Most of the expressions classified as Phrasen by the host are not in any of the specialized dictionaries; nevertheless the defining criteria for FEs doubtless apply to them, as these expressions are very frequently used in a stable form in talk about football. However, many phrases go "unpunished", whereas others, like golfer Bernhard Langer’s extremely original and innovative Leistungssport spielt sich zwischen den Ohren ab (cf. 3.2., (13)) are considered a Phrase. But the aim of the Phrasenschwein is of course not to deliver a systematic linguistic analysis, but to amuse the spectator. The negative evaluation of Phrasen also seems to be motivated by the principle "Variatio delectat", the rhetorical maxim which implies a recommendation to avoid stereotyped language. Stein (1995: 89) calls this the "Mythos vom Formel- oder Floskeldeutsch, oder: Sprache aus der Konservenbüchse"47 and criticizes it as a conservative view of style. As a matter of fact, considering the (almost) inevitable use of prefabricated language (cf. Gülich, 1988/1997), it is this FE-hostile attitude that seems of another age, not the use of fixed expressions. The systematic analysis of authentic corpora would certainly help to establish empirically valid accounts of real-life use of phraseological expressions and thus do away with stereotyped prejudice against certain types of language use.

4 Conclusion

Four types of metalinguistic treatment of FEs have been examined; they have three basic communicative functions:

(a) metalinguistic terms like Spruch identify the expression referred to as commonly used in a stable form;

(b) various types of metalinguistic comment (nonverbal or verbal quotation marks, if I can put it like that, so-called, literally, in plain words) made by the producer of the FE are used to guide the hearer’s interpretation of the expression concerned;

(c) metalinguistic comments serve to (other-)evaluate phraseological expressions, characterizing them, for instance, as too colloquial or even vulgar or else hackneyed.

However, as opposed to journalistic texts, analyzed by Bastian / Hammer (2000), it does not seem that a major function of the metalinguistic comments of FEs is to assure a "better comprehension" of the expressions concerned (cf. Bastian / Hammer 2000: 300).

From a strict conversation analytic perspective only type (c) covers the "treatment" of FEs from an interactive point of view because only in examples (26) and (27) does the co-participant explicitly orientate his (subsequent) activity to his partner’s (prior) phraseological or at least stable activity. In the first two types, the producer of a FE himself comments on it by using a preceding or following metalinguistic evaluation. In Gülich / Kotschi’s (1995) terms these auto-produced meta-discursive comments and/or evaluations can be considered as part of formulation procedures in discourse production. Following a classification and structural description of various types of metalinguistic comments on FEs, interpretations as to their conversational functions have been formulated. Given the (official) absence of specifically oriented co-participant reactions validating these interpretations, they cannot claim more than the status of hypotheses which are, however, the result of detailed analyses of multiple instances of metalinguistic comments.

Whereas metalinguistic other-evaluations of FEs could threaten the speaker’s face as a competent member of the speech community, it seems that speaker-(auto)produced metalinguistic comments play an important role in the constitution of reciprocity between speaker and hearer. By indicating explicitly that he/she does not want to be too direct or categorical, that his turn is not to be taken literally, i.e. by giving directions on how to interpret his turn-at-talk, the speaker demonstrates a high degree of reciprocity. At the same time he takes care of his own face, presenting himself as a competent speaker and distancing himself from formulations which are widely used within the speech community and may thus be considered hackneyed.

A total of 44 instances of metalinguistic treatment of (potential) phraseological expressions were counted in the corpus: 14 of these were commonplaces (7 from football), amounting to 32% of the FEs treated metalinguistically, and 17 metaphorical fixed expressions (39%).48 It is not surprising that the two main types treated metalinguistically were metaphorical expressions and (hackneyed) commonplaces. Non-idiomatic collocations49, e.g. to brush one’s teeth, do not need to be treated metalinguistically; at best they could be evaluated as over-used (stop saying this all the time) or be characterized as a specific speaker’s style.

Of course, not every single idiomatic expression is treated metalinguistically, possibly because there is no conversational need to do so or because some FEs have become so much part of the language that speakers no longer notice their phraseological character50; examples are Gras über etwas wachsen lassen (let the dust settle on sth.), keine Garantie für etwas haben (have no guarantee sth.), Feuer und Flamme sein (be as keen as mustard), in erster Linie (in the first place), einen Schlussstrich ziehen (consider sth. finished), which are used by speakers but remain untreated in the corpus in spite of the images and/or metaphors they contain.