Navigation – Plan du site

Final but, theticality and subjectification

Sylvie Hancil


La particule finale but a été l’objet d’une attention particulière dans de récentes études en anglais américain et en anglais austalien. À présent, il paraît pertinent d’examiner d’autres dialectes de l’anglais. Le but de cet article est d’offrir une étude synchronique de la particule finale but en anglais britannique, pour ses variantes du sud et du nord, à travers l’examen de la partie synchronique orale du British National Corpus et la partie orale du Scottish Corpus of Texts and Speech. En s’appuyant sur les cinq critères qui définissent la catégorie des thétiques, nous montrons que la particule finale but peut être classée comme un thétique avec une structure schématique. L’analyse se focalise aussi sur la façon dont est subjectivisée la particule finale but dans les deux régions.

Haut de page

Texte intégral


1Various studies have recently underlined the importance of dialogicity and interactionality to favour the emergence and the development of linguistic expressions (Waltereit & Detges 2007, Traugott 2010b and Haselow 2014) and this work is situated within this tradition. The use of final particles is quite well-documented in South-Asian languages such as Chinese, Korean and Japanese, as instantiated in (1)-(2):

(1) Korean examples (cited from Kim and Sohn 2015):

  1. cey-ka tayphyokitopwuthak-tulilyekwuha-nuntey
    ‘I wanted to ask you to do the representing prayer, though.’

  2. pang-i nemwu cak-a kuntey
    The room is too small, though.’

(2) Japanese examples (cited from Izutsu and Izutsu 2014):

  1. Sore-tte takai-yo, demo. Its expensive, though. (Common Colloquial)

  2. Syasin tore-n-gana, sikasi. I cant take a picture,though. (Osaka)

2The phenomenon has been relatively little studied in European languages. In English, it is the final particle but that has attracted the attention of researchers such as Mulder & Thompson (2008) and Mulder, Thompson & Williams (2009) in American and Australian English. The purpose is to close the gap in British English and provide a synchronic study of final but in Southern and Northern English, which can be exemplified in (3)-(4):

(3) [Extract from a conversation] He’s only using fibre-glass and they’re not that flammable No, what they’re saying is. Not the flammable, it’s the er indemnity, nobody can breathing it in who’s in Oh, yeah, yeah. They’re frightened like, in years to come, like, they’re gonna try and make a claim against them ‘cos their er, been on the chest, is summat like that. Yeah. I, it’s health and safety, yeah. er. Through the health and safety, but. Eighty four we had this extension put up. Yeah. Sasha come in here. (BNC, KB1)

(4) [Extract of a conversation bewetween a Glaswegian woman and an interviewer]
F646 And I says but, I says, "can you be half a wean?" My father says, "what in God’s name are you talkin about thi- noo [laugh] this time?" I says "Uncle Willie says that Eleanor, they’ve got to be awful careful with her, because she’s only hauf a wean." "In the name o God", he says, "Look, she’s a the wean she’ll ever be. She’s got all her fingers, and all her toes, and all the other bits." He says, "But she was a, she was a twin and she was the one that survived",
F632 mmhm

F646 So this was it, she was only hauf a wean, [laughs] half o a set of twins [laughs]. My father says, "I don’t know what some folk use for brains!" [laughs], ah Go-, but, think about it. (Scottish Corpus)

  • 1 See section 3 for a full description.

3The main differences with American and Australian English final but are : there are no Janus-types1 of but in the two corpora examined and the particle has developed meanings other than the contrastive semantic value, as will be demonstrated in this article.

4Before starting the analysis, data and methodology are presented in section 2. Previous literature on the phenomenon is discussed in section 3. A theoretical background is also provided (section 4). The distribution in the two corpora is delineated in section 5. Relying on the five criteria that define the category of theticals, it is shown in section 6 that final but can be classified as a thetical (Kaltenböck et al. 2011: 861). It is demonstrated that final but is a type of constructional thetical (Kaltenböck et al. 2011: 875) as it has a schematic structure of its own. The discussion is followed by an analysis of the degree of subjectification in the two regions.

1. Data and methodology

5Because it is intended to illustrate the phenomenon of final but in both Southern English and Northern English, the following two corpora were used:

6The spoken demographic section of the British National Corpus (hence BNC-SD) is a corpus of spoken English of about 5 million words, which dates back to the 1990s. A large variety of subjects were examined, extending from broadcast discussions, documentaries, news, classrooms, courtrooms, everyday conversations to lectures in humanities, science, politics and commerce. The corpus was accessed online through Sketch Engine.

7The Scottish Corpus of Texts and Speech is a corpus of almost 5 million words, the spoken section being above 1 million words. Recordings of the spoken texts started in 2000. The corpus includes speakers from a wide range of geographical sites of Scotland. Various types of recordings were conducted: from conversations between mothers and their young children, or a number of interviews with people talking about language issues in Scotland to documents involving people from the Scottish Parliament. Transcriptions of audio recordings were made using Praat software, which enabled them to be time-stamped and synchronised with the recording. The corpus was accessed online.

8To study and analyse the corpora, this author resorted to the help of native speakers representing the dialects used (Southern English and Scottish English).

2. Previous literature

9In analysing the use of final but in American English and Australian English conversations, Mulder and Thompson (2008: 180) argued for the grammaticalization cline shown in (5):

(5) Initial but > Janus-faced but > final but
IU-initial conjunction IU-final discourse particle

10Mulder and Thompson (2008: 195) distinguish six stages on the grammaticization pathway, which are syntactically described as in (6a-f), respectively. Square brackets indicate an intonation unit, and X and Y semantically conflict each other through a contrast or concessive relationship. The initial and Janus 1 buts begin an intonation unit. Janus buts are “between” the initial and final buts, and are called “Janus” because they have properties of both “initial” and final buts, and can be interpreted as either.

(6a) Initial: X [but Y]

(6b) Janus 1: X [but,] Y

(6c) Janus 2: [X but]

(6d) Final 1 (final hanging): [X but]

(6e) Indeterminate: [X/Y but]

(6f) Final 2 (final particle) but): [Y but]

11Mulder and Thompson (2008) show that but follows a grammaticization pathway that starts as an initial but as illustrated in (7a) and becomes Janus-faced but as exemplified in (7b-c) to eventually reach the stage of final but as shown in (7d-f):

(7a) So he got another radio this summer, but of course that got rippled off also. (Initial)

(7b) I dont know what the real story is, but, ... it sounded kinda neat. (Janus 1)

(7c) ...we would charge (H) ...five-hundred fifty dollars on an- on an account, it would be five-hundred dollars, its really kind of switch around but. (H) what...what that would-...I think it would be good for (H) ... the five or six of us, (H) to have Galino down here, (H) can kind of explain what products, ... we can offer from the bank side, ~ Matt needs to know that, and ... and we all need to know that, (H) and then, we can figure out how ~ Matts, ... the products that LCLs gonna offer will plug into that. (Janus 2)

(7d) Wl now Didier - makes his money by going to Atlantic City but -. (Final 1)

(7e) Yep. They are now any way. mum doesnt think so - but, (Indeterminate)

(7f) You sounded funny @@(H) I know. Sounded like an alright person but. (Final 2) (from Izutsu and Izutsu 2014)

12In the initial example (7a), initial but begins an Intonation Unit (IU) and the contrasting value is provided with the preceding sentence (for more details, see Schiffrin 1987 among others).

13In the Janus 1 example (7b), we can say that prosodically it can be assimilated as a final but upon its production, but in reality the speaker immediately gives further contrasting material in the same turn, which transforms it into an introducer. In the Janus 2 example (7c), the speaker continues his turn and there is no evidence that the subsequent material stands in contrast with that preceding the but; it is used in fact as introducing a social action.

14A final particle is a discourse marker that occurs at the end of an interactional unit, whether a turn, a turn unit, or a prosodic unit, and indexes certain pragmatic stances. (Mulder and Thompson 2008: 183). While American English and Australian English have final hanging but, only Australian English has final particle but. In the final 1 example (7d), there is a clear implication left “hanging” such that the clause with but is open to being interpreted as a concession. This but tells the hearer that there is an implication, and invites the hearer to infer what it is and to continue the interaction appropriately given that implication. The example (7e) is an in-between case, where but can be interpreted as a final 1 or a final 2 instance. It is said to be indeterminate. In the final 2 example (7f), the semantically contrastive material is given in the IU ending with the final but particle. The particle closes a construction which conveys the semantically contrasting content.

15Before moving to the analysis of final but in the two corpora, some theoretical background will be provided to account for the thetical nature of the particle.

3. Theoretical background

16Following Kaltenböck, Heine and Kuteva (2011: 850), Discourse Grammar is defined as being composed of all the linguistic resources that are available for constructing spoken or written texts. Sentence Grammar includes parts of speech or constituent types such as sentences, clauses, phrases, words, and morphemes, along with the syntactic and morphological equipment to relate constituents to one another. Thetical Grammar consists of a directory of thetical expressions as well as the capacity of coopting new theticals from Sentence Grammar with a view to organizing discourse.

17Theticals consist of a word, a phrase, a clause, or even a chunk that does not form any syntactic constituent. Most authors agree to present them as “non-syntactic” phenomena that raise a problem of syntactic classification (Burton-Roberts 2005). This definition in fact covers most of the linguistic expressions that are traditionally analysed as Discourse Markers. The class of theticals includes interjections, particles such as though or but when used sentence-finally, certain types of parenthetical comment clauses, such as I believe and as you know, as well as formulae of social exchange such as thanks and I’m sorry, conjuncts (Quirk et al. 1985), or linking adverbials (Biber et al. 1999: 880):

(8) after all, anyway, as it were, besides, however, indeed, in fact, instead, I mean, now, okay, so, then, I think, well, what else, you know, you see (Heine 2013: 1208)

18Three types of theticals can be distinguished. First, there is the type of instantaneous theticals: they are fully compositional, can be formed freely anytime and anywhere, can be inserted in most syntactic slots of a sentence, and quite a few of them are uttered only once:

(9a) And of course during the nineteenth century we can’t hide the fact Egypt became very much a hunting ground for agents on behalf of museums in Europe. (ICE-GB: s2a-052–81)

(9b) ‘In those days,’ Sue admitted, ‘we were heavily in debt.’ (Huddleston and Pullum 2002: 1026–7)

(9c) The therapist’s level tone is bland and neutral — he has, for example, avoided stressing ‘you’. (Huddleston and Pullum 2002: 1355 – cited in Kaltenböck et al. 2011: 875)

  • 2 The term ‘construction’ is used in the sense of Construction Grammar (CG). While this is in accorda (...)
  • 3 The term ‘schematic’ is used as in Construction Grammar (CG). In terms of garden-variety CG (Goldbe (...)

19Second, there is the type of constructional theticals, which form our second category, and are recurrent patterns or constructions2 of theticals, being compositional but having some schematic3 structure and function.

Afterthoughts (e.g., repairs)
Jake, I mean Ian, went home already.
Right-dislocations Why, you have to wonder, do they bother with it? The “peace process”, that is. (The Economist, 28-8-2010: 36)
Phrasal appositives Jake, our boss, told us to stay.
Appositive relative clauses And then you had a theologian talking about the Big Bang <,> which I thought was brilliant. (ICE-GB; Dehé and Kavalova 2007b: 3)
Insubordinated clauses If you could open the window, please.
And-clauses Because on this theory and it’s very deeply held uh good educational news is by definition inadmissible (Kaltenböck, Heine and Kuteva 2011: 856-857)

20Third, there is the type of formulaic theticals (Kaltenböck, Heine and Kuteva 2011: 876). These are non-compositional information units: most of the time their form is invariable. They are usually short expressions, morphosyntactically unanalyzable, tend to be positionally flexible and to express functions that are mostly procedural. They relate to the situation of discourse rather than to sentence syntax, as exemplified below in (11):

Conceptual theticals as it were, for example, if at all, if you will
Formulae of social exchange
Good Morning, hello, please
Sir!, Waiter!
Come on!
boy, damn, fuck, hell, ouch, pst, um, wow

4. Distribution of final but

21Before identifying the type of thetical at stake, let us examine the distribution of the particle in the two corpora.

Table 1: Final linking adverbial but in the two corpora



Spoken Scottish corpus


of words

4,890 m


Total number

of but


(0.56% of the

total number

of words)


(0.95% of the

total number

of words)

Total number

of final but


(0.28% of the

total number of but)


(1.80% of the

total number of but)

22Table 1 gives the distribution of final but in the two corpora. In the BNC-SD, there are 27,667 occurrences of but (0.56% of the total) and 78 tokens of final but (0.28% of all the occurrences of but), hence 16 occurrences of this marker per one million words. By way of contrast, in the spoken Scottish data, the total number of but represents 0.95% of the total number of words, which accounts for a 70% difference compared to the BNC results. So we would expect an increase in the percentage of final but in the same proportions. In fact, there is a 543% increase in the Scottish corpus, compared to the results in the BNC-SD: there are 74 occurrences of final but, hence 1.80% of the total number of but. So, the results are more marked in the Northern corpus.

23The results can be refined by using a sociolinguistic parameter (gender):

Table 2: Distribution of gender in the two corpora (raw figures and %)






47 (8.41%)

31 (6.09%)

Scottish Corpus

48 (24.37%)

26 (15.66%)

24Table 2 displays the distribution of final but in relation to gender in the two corpora. Both in raw figures and in percentage, women use the final particle more than men. The margin is more important in the case of the Scottish corpus (+50%).

5. Final but and theticality

25Final but is a particle, which is assimilated to so-called parenthetical constructions and various extra-clausal units such as vocatives, imperatives, formulae of social exchange, and interjections (Heine, Kaltenböck and Kuteva 2015). The purpose of this section is to examine the nature of the thetical but.

5.1 A prototypical thetical?

26The prototypical properties of theticals (Kaltenböck et al. 2011: 857; Heine et al. 2013) can be stated as follows in (12):

(12a) they are syntactically independent.

(12b) they are set off prosodically from the rest of an utterance.

(12c) their meaning is “non-restrictive”.

(12d) they tend to be positionally mobile.

(12e) Their internal structure is built on the presence of SG but can be elliptic.

27These properties will be tested with final but to see to what extent it can be classified as a thetical, a marker taken out of Sentence Grammar with a view to organizing discourse. Let us review the points one after another for the marker examined.

28It must be admitted that we did not have access to the sound files of the marker but the transcriptions represent a linguistic testimony of some prosodic features. When final but is preceded by a comma, it means that it is separated from the anchor clause and stands on its own, which means it is syntactically independent. Let us see the distibution in the two corpora (Table 3).

Table 3: Distribution of commas before final but in the two corpora (raw figures and %)


But preceded by a comma

But not-preceded by a comma


36 (46.1%)

42 (53.9%)

Scottish Corpus

40 (54.1%)

34 (45.9%)

29The distributions are practically equivalent: in the BNC-SD, there are slightly more markers integrated into the clause (53.9%); in the Scottish Corpus, the distribution is reversed (over 54% of the markers are preceded by a comma).

30It is possible to relate these results to the parameter of gender (Table 4).

Table 4: Distribution of the punctuation before final but in relation to gender (raw figures)


But preceded by a comma

But not-preceded by a comma


Female 21

Male 15

Female 26

Male 16

Scottish Corpus

Female 24

Male 16

Female 24

Male 10

31In both corpora, there are slightly more women (in raw figures) than men who use final but. Whether the marker is preceded by a comma or not, there are always 50% more women than men using the marker. In addition, not only do women use final but more extensively but they also contribute to integrating it more regularly to the anchor clause.

32Final but is said to be “non-restrictive” (Huddleston & Pullum 2002: 1350-62) since it is not semantically part of the clause it is inserted in. It is not semantically linked to the host utterance it is related to.

33Even though theticals tend to be mobile (see Heine 2013), the marker considered is restricted to the final position of an utterance. It can be argued (see section 7) that the postposing of constituents is directly related to the speaker’s discourse strategy, and can be accounted for in interactional terms. Informal conversations involve exchanges between interlocutors, and as a result, they are typically hearer-oriented. The syntactic structure, i.e. the linear order of constituents, is also adapted to these interactional forces. Various studies in discourse analysis have long established that the ideal place for the hearer to manifest his point of view about the speaker’s sentence, namely to index his stance, is at the end of the speaker’s sentence (cf. Pomerantz 1984, inter alia).

34Final but can be elliptic since it does not contribute to the truth conditions of the clause it is in. You can delete the particle without raising a problem for the meaning of the sentence. Consequently, after reviewing the criteria for theticality, it can be said that final but can be classified as a thetical, even though it is not a prototypical thetical. It will be shown in the next section that it can acquire a large variety of meanings, all being related to the discourse situation.

5.2 A schematic structure of final but

35The purpose of this section is to show that final but is a constructional thetical, which has a schematic structure of its own. The starting point of the diachronic development of the marker was its use as a preposition meaning "outside of", as in (13):

(13) butan leodrihte « outside of the law of the land » (Andr. Kmbl. 1357; The Legend of Saint Andrew. 679, cited in the OED)

  • 4 Beal (p. c.) states that final particles are associated with a low flat intonation, which makes the (...)

36In the case of final but, which has the semantic structure X-but-Y, X is the preceding discourse and Y is implicitly present.4 It will be demonstrated that the various meanings attached to the final particle but can be derived from its fundamental spatial value.

37In example (13), given the original meaning of but, a clear division is found between two spaces: if s stands for space, there are inside s and outside s, where s is already a case of metaphorical extension of a local notion to a legal one (literally “outside of the law of the land”). In addition to this basic local meaning we have also found since Old English an extension to a more abstract meaning ‘except’ (see 14a). Here we find the basic asymmetry expressed by the preposition but and the division into two complementary spaces in a more abstract shape. See also (14b):

(14a) You were assailed by a warm stink that all but strangled you. (G. de Poncins & L. Galantière Kabloona (1942) i. iv. 101)

(14b) I want nothing but the best for my children.

38The omission of the negative quantifier results in a focus particle meaning ‘only’.

39In the development of butan from a preposition to a conjunction, two steps can be distinguished. In (15) but takes a clause introduced by a subordinating conjunction as its complement instead of a noun phrase. Its meaning, however, is still very much that of a restrictor for the quantifier.

(15) I think it scarcely possible, but that some of the enlarged minds..will coincide with me. (Mary Wollstonecraft, A vindication of the rights of Woman 1792, Ded. p. v)

40When but is used with sentential complements without subordinating conjunctions this marker then can be analysed as a coordinating conjunction (p but q). At that stage the two spaces are argumentative ones and the asymmetry resides in the fact that the two clauses support different or opposite inferences (cf. Anscombre & Ducrot, 1977).

(16a) p & q

(16b) p is an argument for the inference ‘r’
q justifies an opposite inference ‘non-r’

41This asymmetry is exemplied in (17):

(17) John cut his finger but he did not cry.

42where p is asserted (‘John cut his finger’) and p suggests the inference ‘r’ (‘John will cry’), while q is asserted (‘he did not cry’) and q justifies non-r (‘He will not cry’).

43The development of the conjunction into a final particle could be the result of (a) reanalysis or of (b) ellipsis:

  1. p but q > [p but] q > change of order p, q but.

  2. p but q > q but…

  • 5 But and though can be interchangeable. Their use is then a matter of regional variation.

44At this stage but can be paraphrased by ‘though’5 and just rejects a plausible inference:

(18) It is a rainy day but it is warm > It is warm but.

45The rejected inference is “It is a cold day.”

46The concessive meaning persists with a higher proportion in the BNC-SD (29 occurences, 37.2%) than in the Northern corpus: the occurrences are rare in the spoken Scottish corpus (1 occurrence, 1.4%). The marker can be paraphrased with though in the examples (19) – (20):

(19) I’ve just got to tell her a secret. Remember I’ve got to see. Yeah, at two fifteen ain’t it? That means at two fifteen. Two fifteen is quarter past two. Anyway. I’m thirteen by the way, and Jo’s, how old are you Jo? fourteen? Jo’s fourteen. You see, and she didn’t do anything for her birthday, the stupid cow. I wanna do something. When? I was gonna do something last Sunday, but. Do something on Saturday. Do something on Saturday, yeah. No, I don’t want to Saturday. Do something on Monday. I prefer Sunday. Yeah, go on then. (BNC, KNY)

(20) F643: but she’s very good in intuiting whether you actually need that or whether you need, ehm, a dose of herbal medicine or whether you just need to calm down.
M608: mm Yeah.
F641: mmhm
F643: So she doesn’t always give you it. So that, that would be one Iwould like to learn eventually,
M608: mmhm
F643: erm
F641: mmhm
F643: because I think I’m drawn to it obviously from my design background
and the colour side
M608: mm
F643: all of the courses I’ve checked out are hideously expensive, ehm.
(Scottish corpus)

47In the BNC example, the speaker explains what she did on second thought and marks, with the use of final but, a contrast between what she intended to do for her birthday and what she actually did. In the Scottish example, the speaker explains that she is attracted to the colour therapy because of the years she spent doing design and makes a contrast between design and colour, using final but.

48The final particle can also acquire an anaphoric/cataphoric value, which can be paraphrased by “you know”. The preceding text (“anaphoric value”, as in [21a, b]) or coming text (“cataphoric value”, as in [22a, b]) can be looked into to identify the second proposition. The indication of a dissociative association is still present in but: the speaker goes “out” of the proposition to look for a referent and makes a link with this referent. No mention of the locator is explicitly made and it ends up being under­stood in light of the preceding context; hence the value of anaphor.

49Final but takes an anaphoric value in very few occurrences in the BNC-SD (4 occurrences, 5.1% of the total) and in a significant number of examples in the Northern corpus: 20 occurrences (27%) in the Scottish corpus. Examples (21a-b) show the anaphoric value of final but:

(21a) SP:PS066 And Robin and L-- and erm (pause) Carol doing that er cabbage (pause) cabbage
SP:PS065 Oh that revolting
SP:PS066 That was horrible.
SP:PS065 That was terrible. (pause) I remember Jill though. I don’t remember the mix but mm I don’t know that, I vaguely remember now, yes but
SP:PS066 Oh they were all there (SP:PS065) Mm.
SP:PS066 because we’d gone out, I think it was the christmas shopping we were doing on
our own.
SP:PS065 Was it?
SP:PS066 Mm.
SP:PS065 Mm. Yes I’ve got a hazy recollection
but. Mm. (pause) Jill’ll remember. But she was a lot younger surely, Jill? And she’d be a bit
SP:PS066 Oh she’d be young, erm
SP:PS065 Younger than them I mean. (BNC, KBP)

(21b) F1053 When would ye use skelp?
M1000 I would do.
F1053 You would?
M1000 Aye.
F1053 An when when would you use it?
M1000 Mm now eh [exhale] Probably ah. Aye, oh aye.
M999 /Well ye gie a beast a skelp,
M1000 Nae bother. [laugh]
M999 that’s it, that’s right, no no, we didnae get skelped, no no, no. We got bashed. [laugh] Gied a bash. Eh eh but that’s nae a eh a kinda localism
F1053 Kay, would you have been skelped?
F1001 Oh definitely, definitely skelped. (Scottish corpus)

50In example (21a), the speaker discusses his vague memories of Christmas shopping (hazy recollection) and points to the preceding context I vaguely remember. In (21b), the speaker discusses the meaning and use of “skelp” and points to the idiosyncratic use of “skelp” by saying it is a “nae a kinda localism”, as underlined by final but.

51The percentage of examples with a cataphoric interpretation is three times as high as that for anaphoric meaning (15.4%, 12 occurrences) in the BNC-SD corpus, whereas in the Scottish corpus, the percentage is a bit higher (17.6%, 13 occurrences) for the cataphoric meaning. Examples (22a-b) illustrate the cataphoric value of final but:

(22a) Oh yes, I’d never get a professional job on that heavy cloth and all the so at some point that’ll. So, is it quite big though the suit? Oh yes, thirty three and thirty eight, he tried thirty six. That’s not fitting. He’s, he’s not. I wasn’t too sure at first if it’ll fit him, but when he put the trousers on, when he put the trousers on first, I thought no, no way, but then when pinned them, you know, they I thought, I thought, but. Told him to go to grey though, cos grey won’t date, green will. No, that, nice with a black jumper. (BNC, KBF)

(22b) F1037 of the best Polys they didn’t need to change them. And she did English, erm and she did have friends who were at Oxford colleges, doing English, and funnily enough, erm they reckoned that erm, thi- this probably is politically incorrect, I’ll have to cut this, er but they reckoned that Oxford Brookes was offering a better English course than Oxford Uni, because they hadn’t really changed the curriculum at the University, er at Oxford University, for ages and ages, and she was doing all sorts of things, like Caribbean poetry and really kind of funky stuff, and they were they were still doing the same old things. And erm
F1038 Right.
F1037 and she did that with Catering. She had a wonderful time, great place to be at, Oxford, and she did have friends at the college too,
F1038 Yeah.
F1037 she’s not hugely academic, erm but erm er intelligent, but I think she doesn’t really
like having her nose in a book all the time, so it was a great place for her. Er,
F1038 Yeah. (Scottish corpus)

52In example (22a), the speaker discusses the best way of dressing up for a professional job and he makes a few remarks about the colour of the trousers. With the final particle but, he points to upcoming information, which he specifies in the next sentence with the adjective grey. In example (22b), the speaker remembers how life was at Oxford colleges and points, through the use of final but, to upcoming information about her friend’s character “she’s not hugely academic”.

53The final particle can also take on an intensifying value, paraphraseable by « really ». The marking of emphasis reaches almost 18% (14 occurrences) in the BNC-SD. It is slightly less represented in the Scottish corpus (16.2%, 12 occurrences). Examples (23a-b) delineate the intensifying value of final but:

(23a) Yeah, dry is dry autumn too, you see. Yes. Yeah. Yes. They didn’t really get a hold did they? They didn’t get a hold, didn’t get a start. Really? Aha. See an a, I wouldn’t like other people keep using the water cos they very careful on using the water Yes. but if I’d have give it, had to give them some water they’d have done much better but I, like everybody else we kept hoping it were gonna rain, but! That’s right. Yeah. (BNC, KC0)

(23b) F1009 Scaff, I have no idea where that comes from. Scaff?
M1055 Wo-wo- is that used in Skye?
F1009 Mmhm [inhale] Well apparently it’s used in Portree High School. Yes. [laugh] Now it certainly wasn’t in in my day [?]we never[/?]
M1008 Mmhm. Mmhm.
F1009 not in in Alister’s
M1007 Completely at a loss with this one. (Scottish corpus)

54In the English example, the speaker talks about how dry autumn was and how precious water was; she finishes up commenting about how they kept hoping for rain. The final particle but is followed by an exclamation mark and a commentary of approval “that’s right”, which shows the intensifying value of final but. In the Scottish example (23b), the two speakers are looking for the meaning of the word « "scaff" ». Speaker F1009 remarks that it was not used in his days and insists that it was not used in one of his friends’s day either (not in in Alister’s) by adding the final particle but.

  • 6 There is no trace of semantic meaning in this use of the particle. This use represents the final st (...)

55The final particle may eventually lose all its content and become a mere filler, an expression of a phatic value as a result of erosion. The marker is devoid of any semantic value and is used purely to fill in a gap in the conversation, allowing the conversation to keep going without any interruption. Even though the marker is devoid of information, it is used as a guidance for the hearer to keep the conversation going. The use of the marker with a phatic value represents 24.3% of the occurrences (17) in the BNC-SD, namely the highest percentage among the various meanings of final but in the BNC-SD. By contrast, it is in the Scottish corpus that one can see the highest percentage (28 occurrences, 37.8%) for the phatic value.6

(24a) Yeah. Well, yes. I know about cars Marion. C reg C reg Fiesta, how much do you reckon a C reg Fiesta costs? Can we see it? No. I’m just gonna buy it. Well I’m not, my dad and mum are, just don’t be impulsive. It’s only one thousand five hundred though, that’s alright. Ah ah! Yeah but. Yeah. No problem. Couldn’t I mate? Yeah. I’ll buy a couple. (BNC, KP5)

(24b) M811 No, there’s usually like, the Leeds ones though, like, I’ve only
F812 [laugh]
M811 ever been twice and only for the day just because there’s only really been one day, //well my brother lives in
F813 Yeah.
M811 Leeds so we just sort of go
F814 Uh-huh
M811 occasionally but it’s usually quite a mix, there’s usually one day that’s more like, sort of rock and metal stuff and me and my brother go to the one that’s
F813 And then they have a t-, yeah.
M811 a bit more indie, one that’s a bit more somethin else,
F812 I’ve honestly can never say I’ve ever kinda got into the whole proper concert.
F813 Yeah. (Scottish corpus)

56In example (24a), the speaker is wondering how much a Fiesta model car costs. He manages to find the right price (It’s only one thousand five hundred) and adds the particle but in final position to keep the conversation going. In example (24b), the speaker discusses with his friend the type of music they listen to in Leeds. He identifies the music as being “a bit more indie” and complements his comment by adding final particle but to show it is the end of his turn.

57The various meanings of final but can be summarized in Table 5:

Table 5: Meanings of final but in the two corpora (raw figures and normalized values)

Corpus/Nb of occurrences



Scottish Corpus


29 (9.04)

1 (0.037)


4 (0.04)

20 (6.37)


12 (2.92)

13 (4.04)


14 (3.64)

12 (3.70)


19 (5.44)

28 (9.04)

58In the Scottish corpus, the phatic value leads as the most often represented, with the contrasting value being the least represented. This could be explained by the fact that the semantic value has lost its original value through erosion, this original value remaining present in other categories. By way of contrast, in the BNC-SD, the contrastive meaning leads the way, followed by the phatic value; this could be accounted for by the fact that the use of final but is relatively new, so it retains its original value.

59Consequently, after examining the various meanings of the final but in relation to its original spatial meaning, it is possible to posit that final but has a schematic structure of its own X but Y and is a constructional thetical.

6. Final but and subjectification

60Since final but is used in interaction, it would be worth studying the degree of subjectification of the particle in the two regions, a process that is a strong correlate for grammaticalization, extending from non-subjective to intersubjective vua subjective (Traugott and Dasher 2002). Based on extensive data analysis, Ford and Thompson (1996) show that turn taking usually occurs at what they call “Complex Transition Relevance Places”, which are the locus of the interaction of pragmatic, syntactic and intonational completion points:

(25) Intonation and pragmatic completion points select from the syntactic completion points to form what we will call “Complex Transition Relevance Places”. The term “turn unit”, then, will refer to a unit which is characterized by ending at a CTRP. (Ford and Thompson 1996: 154)

61In their study, there is a high proportion of speaker change that occurs at CTRPs (71%) but it is possible to find a significant number of occurrences of speaker change in non-CTRPs, where syntactic completion points do not coincide with intonational completion points (47%) and for which speaker change can only be explained in interactional terms (Ford and Thompson 1996: 159-171): the speaker uses final markers in non-CTRPs to guide the hearer in a certain direction of the conversation that he wants to adopt, thereby helping the hearer to infer the appropriate implicature. That is the conclusion (Ford and Thompson 1996: 158) we have tested against our data.

62In our BNC data (see Table 6), final particles do not generate speaker change, whether it be in CTRPs or in non-CTRPs. Interestingly, almost all of the occurrences in the BNC with final particles are represented by the Southern English regions. In the Scottish English data, however, speaker change can occur after a CTRP or after a non-CTRP; the distribution of the occurrences of final but tends to slightly favour speaker change after a CTRP. Thus, the results show that the number of speaker changes increases the more we move to the North of Britain. The findings confirm, too, that the increasing use of final markers in non-CTRPs is directly related to the increasing use of interactional motivations.

Table 6: Speaker change in the BNC-SD and the Scottish Corpus (raw figures and %)


Scottish Corpus

Speaker change after a CTRP





Speaker change after a non-CTRP





63Let us note that there is not a single occurrence of final but in the BNC-SD followed by speaker change. However, the final but data in the Scottish corpus are systematically followed by speaker change.

64The use of final constructions by the speaker contributes to maintaining the preferred organization of conversations by encouraging an agreement from the hearer. As can be expected given its higher degree of subjectification, it is more frequent in the Scottish English corpus. Only 2 examples denoting disagreement were identified in the Scottish data, given here as (26) and (27):

(26) M 1004 : To dog off, dog off.
F 1054 : Yeah, you wouldnae say "skip" or "skive"?
M 1004 : Not with school, [inaudible].
F 11005 : It’s usually
F 1006 : I’ve heard that used too, I think,
M 1004 : Ah we
didnae [inaudible].
F 1005 : Dog school. [laugh]. No that we ever done that. [laugh]. (Scottish Corpus, Inverurie T110)

(27) M 999 Ill-trickit, would you use that word, Dad? No.
M1000 Nae really, nuh, no. No, no.
F1001 I would use that
M1002 It’s
nae ane I would use much like. (Scottish Corpus, Inverurie T110)

65Let us now examine the explicit marking of agreement of the hearer with the speaker’s speech. The acknowledgement is frequently indicated by the use of the adverbs yes/yeah, as indicated in (28):

(28) M734 er, and you come off, it’s gonna be a whole lot worse than doing twenty and coming off, though I’ve got more safety equipment, er on when I’m, er, on the motorbike, er
F718 mmhm
M734 But it’s crashing hasn’t really come into my mind It [laugh] Well, kind of, Mum always, kind of, worries,
M734 I’m much more "what will be will be". If I’m gonna crash, I’m gonna crash. It kind of, life’s full of turns.
(Scottish Corpus, Conversation 01, T60)

66In other cases the agreement is more indirectly expressed by the interactional, repetitive use of the adverbs mm-mm and uh-huh, or even the use of a question tag (cf. (29) and (30) respectively):

(29) F810: Like th- they had great plans like to get a flat together and stuff in St Andrews, which now isn’t happening, so I don’t know really where she wants to go, but.
F810: I don’t know why she’d want to go to St Andrews.
F809: Yeah, that.
F810: It’s kind of quiet. And a little dull. (Scottish Corpus, conversation 11, T105)

(30) M819: fortune. That that that that that place would
M818: I think.
M819: die after twenty minutes and, it, I mean, you’re talkin places in Londo- London are cheaper than this.
M818: Really, aye?
M819: Aye. Cause cause, well it’s the competition.
M818: I’ve I’ve only been in it a couple of times,
Isn’t it? Cause I mean that that noodle bar place, it’s right next to the Garage, it doesn’t have to worry about business.
M818: Aye, it’ll be dead during the day probably, but [inhale] I think it’s
M819: It’s twenty-four hour is it not? (Scottish Corpus, conversation 14, T57)

67The Scottish data also comprise markers of the hearer’s attitude in direct relation to the speaker’s turn, as illustrated in (31):

(31) F834: So [CENSORED: forename] fit do you want to dae when you’re elder?
F835: I’ve got a lot of things I want to dae but I’m nae sure yet. [tut] I’m thinkin I want to work on a cruise ship an do some beauty therapy,
Wow! That’s so cool!
F835: because it, you see I figured you don’t have to pay like, for like tax or anything [laugh], got your food an your like bed free //[inaudible] bed [laugh] (Scottish corpus, conversation 20, T86)

  • 7 We can say that the final particle has a mixed syntactic behaviour if the particle can be preceded (...)

68These results reveal something about different degrees of subjectification in the different regions. In Southern England, the use of final but could be said to emerge as an instance of subjectification on a pragmatic basis only. Syntactically, but has a mixed syntactic behaviour7. The interactional forces do not lead to speaker change; so, final particles in Southern English cannot be considered as being subjectified. In Scottish English, on the other hand, the data show a more advanced degree of subjectification, since speaker change occurs in all the instances except in 2 occurrences of final but. The syntactic behaviour of the markers is also mixed, since they can both be assimilated as orphans or IP-adjuncts, so the syntactic grammaticalization is on its way.

69Let us see how the final particle but can be placed on Traugott and Dasher’s (2002) subjectification cline, a cline extending from non-subjective towards intersubjective via subjective. In accordance with Grice’s Maxim of Manner “be brief”, hearers are invited to make the inference that more (information) is conveyed in an utterance than simply literal information. The meaning of final but is understood in procedural terms (Blakemore 1992). The hearer is given an interpretive cue as to the intended inference she is supposed to draw. In all the observed cases, the speaker guides the hearer’s interpretation of the previous utterance and encourages her/him to make the necessary adjustments to achieve the proper interpretation of the utterance. In this sense, final but can be said to be intersubjective since it is clearly hearer-orientated. Other values can be superimposed, as shown in Table 7:

Table 7: Distribution of the semantic values in relation to the subjectivity scale (Traugott & Dasher 2002)

Semantic value of final but

Subjectivity scale


Metalinguistic + Intersubjective


Subjective + Intersubjective


Metalinguistic + Intersubjective


Subjective + Intersubjective



70When final but acquires a contrastive meaning, it has a metalinguistic value: the utterance in which final but appears is compared semantically to the previous utterance and the comparison is linguistically expressed in terms of contrast. Likewise, when final but is described with a cataphoric or anaphoric value, it can be interpreted metalinguistically since it modifies the linguistic expression of the utterance by creating a special link with the following or previous utterance. Besides, in examples expressing doubt or emphasis, the marker has an epistemic, and therefore, subjective meaning since the speaker uses it to evaluate the degree of validity of the situation and express a personal opinion. In the case of its use as a filler, no other meaning is added to the intersubjective interpretation.

71Thus, the switch from initial but to final but illustrates a movement from typically non-subjective meanings toward primarily intersubjective values, along with a change in scope, going from scope over whole propositions to scope over segments of discourse, which is in line with Traugott and Dasher’s principles of grammaticalization, illustrating an increase in subjectification in the process of grammaticalization.


72This synchronic study of the final particle but in Southern English and Scottish English complements the recent investigations of the particle in American English and Australian English (Mulder and Thompson 2008; Mulder, Thompson & Williams 2009) and contributes to a better understanding of the regional distribution of the various meanings attached to the particle in Britain.

73It was shown that the final particle but in Southern English and Scottish English was a constructional thetical, albeit not a prototypical thetical. Besides, the particle has acquired a large range of meanings in the two corpora examined, compared to the constrastive meaning systematically observed in American and Australian English. These meanings exhibit the textual cohesion and the interpersonal coherence of the marker under consideration. It was also demonstrated that Scottish English final but was more subjectified than Southern English final but on the cline of subjectification elaborated by Traugott & Dasher (2002).

74Besides, this study on the final particle but partakes of the debate on whether final particles belong to grammar or not and challenges the traditional notion of grammar. In investigating the use of final particles, one understands the necessity of enlarging the definition of grammar along the terms developed by Traugott (2003) by encompassing not only phonology, morphology and vericonditional semantics, but also elements that exist in the interaction between speaker and co-speaker(s) and that take into consideration discourse management, information process and the regulation of interpersonal relations between participants.

75More generally, this analysis on the final particle but contibutes to enriching the debate between system- and usage-based approaches and raises new discussions on final particles, a linguistic phenomenon that sheds light on theories of grammar, syntactic theory, cognitive linguistics, discourse analysis, conversational analysis, discourse-pragmatic variation and grammaticalization.

Haut de page


Anscombre, Jean Claude & Oswald Ducrot. “Deux mais en français ?”. Lingua 43 (1977): 23-40.

Averintseva-Klisch, M. “German Right Dislocation and Afterthought in Discourse.” In C. Sassen & A. Benz, P. Kühnlein (eds.), Constraints in Discourse. Amsterdam: John Benjamins, 2008. 225-247.

Biber, Douglas, Johansson, Stig, Leech, Geoffrey, Conrad, Susan and Edward Finegan. Longman Student Grammar of Spoken and Written English. Harlow: Longman, [1999]2002.

Ford, Cecilia & Sandra Thompson. “Interactional Units in Conversation: Syntactic, Intonational, and Pragmatic Resources for the Management of Turns.” E. Ochs et al. (eds.), Interaction and Grammar, Studies in interactional Sociolinguistics, Vol. 13. Cambridge: Cambridge UP, 1996. 134-184.

Haegeman, Liliane. “Parenthetical Adverbials: The Radical Orphan Approach.” In S. Chiba et al. (eds.), Aspects of Modern English Linguistics: Papers Presented to Masatomo Ukaji on His 60th Birthday. Tokyo: Kaitakusha, 1991. 232-254.

Haiman, John. Natural Syntax: Iconicity and Erosion. Cambridge Studies in Linguistics 44. Cambridge: Cambridge UP, 1985.

Hancil, Sylvie. “The Final Particle But in British English: an Instance of Cooptation and Grammaticalization at Work.” In Sylvie Hancil & Ekkehard König (eds.), Grammaticalization: Theory and Data. Amsterdam: Benjamins, 2014.

Haselow, Alexander. “Sequentiality in Dialogue as a Trigger for Grammaticalization.” In S. Hancil & E. König (eds.), Grammaticalization Theory and Data. Amsterdam: Benjamins, 2014.

Heine, Bernd. “On Discourse Markers: Grammaticalization, Pragmaticalization, or Something Else?”. Linguistics 51-6 (2013): 1205-1247.

Heine, Bernd, Kaltenbböck, Kuteva, Tania and Haiping Long. “An Outline of Discourse Grammar." In Shannon Bischoff & Carmen Jany (eds.), Functional Approaches to Grammar. Berlin: De Gruyter, 2013. 155-206.

Heine, Bernd, Gunther Kaltenböck, and Tania Kuteva. “Some Observations on the Evolution of Utterance-final Particles.” In Sylvie Hancil, Alexander Haselow and Margje Post, (eds.) Final Particles. Trends in Linguistics, Studies and Monographs. Berlin: De Gruyter, 2015. 111-140.

Himmelmann, Nikolaus. “Lexicalization and Grammaticization: Opposite or Orthogonal?” In Walter Bisang, Nikolaus Himmelmann & Björn Wieme, (eds.) What Makes Grammaticalization? A Look from Its Fringes and Its Components. Berlin: De Gruyter, 2004. 21–42.

Hopper, P. “Emergent Grammar”. Berkeley Linguistic Society 13 (1987): 139-157.

Huddleston, Rodney and Geoffrey K. Pullum. The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press, 2002.

Izutsu, Mitsuko Narita and Katsunobu Izutsu. “Truncation and backshift: two Pathways to Sentence-final Coordinating Conjunctions.” Journal of Historical Pragmatics 15-1 (2014): 62-92.

Kaltenböck, Gunther et al. “On Thetical Grammar.” Studies in Language vol. 35, n. 4 (2011): 852-897.

Kim, Stephanie Hyeri and Sung-Ock Sohn. “Grammar As an Emergent Response to Interactive Needs: A Study of Final Kuntey ‘But’ in Korean Conversation.” Journal of Pragmatics, vol. 83 (2015): 73-90.

Lehmann, Christian. Thoughts on Grammaticalization. Lincom: Munich, [1982] 1995.

Lehmann, Christian. “Theory and Method in Grammaticalization.” Zeitschrift für germanistische Linguistik 32 (2004): 152–187.

Mulder, Jean & Sandra A. Thompson. “The Grammaticization of But as a Final Particle in English Conversation.” In Ritva Laury (ed.), Crosslinguistic Studies of Clause Combining. The Multifunctionality of Conjunctions. Amsterdam: Benjamins, 2008. 179-204.

Mulder, Jean, Thompson, Sandra A. & Cara Penry Williams. “Final But in Australian English Conversation.” In Pam Peters, Peter Collins & Adam Smith (eds.), Comparative Studies in Australian and New Zealand English. Amsterdam: Benjamins, 2009. 339-359.

Pomerantz, Anita. “Agreeing and Disagreeing with Assessments: Some Features Found in Preferred/Dispreferred Turn Shapes.” In J. M. Atkinson & J. Heritage, (eds.), Structures of Social Action, Studies in Conversation Analysis. Cambridge: Cambridge University Press, 1984. 57-101.

Quirk, Randolph, Greenbaum, Sidney, Leech, Geoffrey and Jan Svartvik. A Comprehensive Grammar of The English Language. New York: Longman, 1985.

Traugott, Elizabeth. “Grammaticalization.” In Silvia Luraghi & Vit Bubenik (eds.), Continuum Companion to Historical Linguistics. London: Continuum Press, 2010a. 269–283.

Traugott, Elizabeth. “Dialogic contexts as motivations for syntactic change.” In Robert A. Cloutier, Anne Marie Hamilton-Brehm, and William Kretzschmar (eds.), Variation and Change in English Grammar and Lexicon. Berlin: De Gruyter Mouton, 2010b. 11-27.

Traugott, Elizabeth and Richard Dasher. Regularity in Semantic Change. Cambridge: Cambridge University Press, 2002.

Waltereit, Richard & Ulrich Detges. “Different Functions, Different Histories. Modal Particles and Discourse Markers from a Diachronic Point of View.” Catalan Journal of Linguistics 6 (2007): 61-80.

Haut de page


1 See section 3 for a full description.

2 The term ‘construction’ is used in the sense of Construction Grammar (CG). While this is in accordance with CG wisdom, Kaltenböck, Heine and Kuteva’s use of "constructional" in "constructional thetical" is not entirely so since for adherents of CG, anyhing is a construction, including a word or a particle.

3 The term ‘schematic’ is used as in Construction Grammar (CG). In terms of garden-variety CG (Goldberg 1995: 2006), constructional theticals can be defined as partially substantive constructions. Take comment clause (I think, you know) constructions as an example of a constructional thetical: the schematic part would be something like [subject + mental process verb] whereas the remainder of the utterance is substantive, that is, it is neither structurally nor lexically filled.

4 Beal (p. c.) states that final particles are associated with a low flat intonation, which makes them final.

5 But and though can be interchangeable. Their use is then a matter of regional variation.

6 There is no trace of semantic meaning in this use of the particle. This use represents the final stage of grammaticalization with the extreme erosion of the meaning.

7 We can say that the final particle has a mixed syntactic behaviour if the particle can be preceded by a comma in some instances (and therefore is not integrated into the previous sentence) while not preceded by a comma in others (and is therefore fully integrated into the previous sentence).

Haut de page

Pour citer cet article

Référence électronique

Sylvie Hancil, « Final but, theticality and subjectification », Anglophonia [En ligne], 22 | 2017, mis en ligne le 22 mars 2017, consulté le 28 juin 2017. URL : ; DOI : 10.4000/anglophonia.1043

Haut de page


Sylvie Hancil

University of Rouen

Articles du même auteur

Haut de page

Droits d’auteur

Licence Creative Commons
Anglophonia – French Journal of English Linguistics est mis à disposition selon les termes de la licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Pas de Modification 4.0 International.

Haut de page
  • Logo Presses universitaires du Midi