Download presentation
Presentation is loading. Please wait.
Published byJeremy Miles Modified over 8 years ago
1
Welsh-English code-switching: a clause-based analysis Margaret Deuchar, Dirk Bury, Elen Robert, Peredur Davies & Jonathan Stammers
2
Goals of this talk To argue that the notion “bilingual clause” is potentially problematic To argue that the notion “bilingual clause” is potentially problematic To demonstrate this with reference to complex clauses To demonstrate this with reference to complex clauses To propose an analysis using “clausal units” - which are not necessarily constituents To propose an analysis using “clausal units” - which are not necessarily constituents To report on the results of assigning (a) linguality and (b) a matrix language using clausal units in an analysis of our Welsh-English code-switching data To report on the results of assigning (a) linguality and (b) a matrix language using clausal units in an analysis of our Welsh-English code-switching data
3
What is a bilingual clause? A clause that contains words from more than one language Example: [ CP maeoonstand-by ] [ CP maeoonstand-by ] isheonstand-by He’s on stand-by.
4
Complex clauses A complex clause is a clause or CP that contains a clause: [ CP1 a b c [ CP2 d e f] ] Within CP 1 (a complex clause), we can distinguish: - [ CP2 d e f] : the subordinate clause - a b c : the “residue” of CP 1
5
Monolingual complex clause Example: [ CP1 so wnes i ddeud [ CP2 (ba)swn I so did I say would I so did I say would I mynddyddSul] ] godaySunday [ CP1 So I said [ CP2 I’d go on Sunday] ]
6
4 patterns of complex bilingual clauses 1. [ CP1 L1 L2 [ CP2 L1 L2 ] ] 2. [ CP1 L1 L2 [ CP2 L1 ] ] 3. [ CP1 L1 [ CP2 L2 ] ] 4. [ CP1 L1 [ CP2 L1 L2 ] ]
7
Pattern 1: [ CP1 L1 L2 [ CP2 L1 L2 ] ] Example: [ CP1 ondmaeomorcheesy but isitsocheesy but isitsocheesy [ CP2 mae (y)nfunny ]yndy ] [ CP2 mae (y)nfunny ]yndy ] is PRTfunnyTAG is PRTfunnyTAG But it’s so cheesy it’s funny, isn’t it.
8
Pattern 2: [ CP1 L1 L2 [ CP2 L1 ] ] Example: [ CP1 hynna (y)dy ’r exam [ CP2 dwi ’n there isDETexam am-I PRT gorod eistedd (y)fory] ] have-to.NONFINsit.NONFINtomorrow That’s the exam I have to sit tomorrow.
9
Pattern 3: [ CP1 L1 [ CP2 L2 ] ] Example: [ CP1 wnes iddeud [ CP2 ohd(o)you did Isay ohdoyou did Isay ohdoyou fancyapintonWednesday?] ] fancyapintonWednesday I said “Oh, do you fancy a pint on Wednesday?”
10
Pattern 4: [ CP1 L1 [ CP2 L1 L2 ] ] [ CP1 maehijustynmeddwl[ CP2 bod isshejustPRTthink be isshejustPRTthink be (y)naquick-fixesi bob dimyeah] ] therequick-fixes to every thingyeah She just thinks there are quick-fixes for everything, yeah.
11
Why are complex bilingual clauses problematic? Complex bilingual clauses create problems for: counts of the proportion of bilingual clauses in a given corpus identification of matrix language (as in Matrix Language Frame model) on a clause basis
12
Counting bilingual complex clauses Pattern 4: [ CP1 maehijustynmeddwl[ CP2 bod isshejustPRTthink be isshejustPRTthink be (y)naquick-fixesi bob dimyeah] ] therequick-fixes to every thingyeah She just thinks there are quick-fixes for everything, yeah. #4' [ CPl L 1 [ CP2 L 1 [ CP3 L 1 [ CP4 L 1 L 2 ] ] ] ] #4' [ CPl L 1 [ CP2 L 1 [ CP3 L 1 [ CP4 L 1 L 2 ] ] ] ]
13
Matrix Language Frame model Unit of analysis is CP: “syntactic structure expressing the predicate-argument structure of a clause” (Myers-Scotton 2002:54) A bilingual CP is one that contains bilingual constituents (Myers-Scotton 2002:56), including other CPs Hypothesis (Myers-Scotton 2002:66): “the Matrix Language does not change at all within the single bilingual CP”
14
Predictions of the MLF Morpheme Order Principle In bilingual constituents […], surface word (and morpheme) order will be that of the Matrix Language. System Morpheme Principle In bilingual constituents, all system morphemes which have grammatical relations external to their head constituents will come from the Matrix Language. (Adapted from Myers-Scotton 2002:59)
15
Problem for hypothesis that ML doesn’t change Pattern 3 examples contain word order patterns and agreement morphology from two languages: Pattern 3 examples contain word order patterns and agreement morphology from two languages: [ CP1 wnes iddeud [ CP2 ohd(o)you did Isay ohdoyou did Isay ohdoyou fancyapintonWednesday? ] ] fancyapintonWednesday I said “oh, d’you fancy a pint on Wednesday?”
16
Solution: the clausal unit The basic unit of analysis must include : simple clauses complex clauses that contain a subordinate clause but without the subordinate clause (‘clause residues’) We call the basic unit of analysis clausal unit. Note: Clause residues are not constituents.
17
Illustrative analysis applying MLF to Welsh-English data Data from AHRC-funded project at Bangor (transcribed using LIDES system) Sample of ca. 70 minutes analysed Data from 3 conversations between 2 speakers Total of 2429 clausal units analysed Total of 1171 clausal units with finite verbs Linguality and ML identified
18
Identifying linguality 1. Identify all clausal units (including both simple clauses and clause residues) 2. Code each clausal unit as monolingual (Welsh or English), bilingual or either 3. Calculate proportion of bilingual vs monolingual clausal units in sample
19
Identifying matrix language 1. Identify all clausal units (including both simple clauses and clause residues) 2. Identify ML of each clausal unit on basis of subject-verb agreement & word order as Welsh, English, either or dichotomous 3. Calculate proportion of bilingual and monolingual clauses in which ML can be identified
20
Linguality: all clausal units Number% Welsh170270% English1496% Bilingual35114% Either2279% Total2429
21
Linguality: clausal units containing verbs Number% Welsh126879% English624% Bilingual28318% Either00% Total1613
22
Linguality: clausal units containing finite verbs Number% Welsh90577% English565% Bilingual21018% Either00% Total1171
23
ML of all clausal units Number% Welsh120850% English904% Either1094% Dichotomous1 0%0%0%0% n/a 1022 42% Total2430
24
ML of clausal units containing verbs Number% Welsh115171% English553% Either413% Dichotomous1 0%0%0%0% n/a365 23% Total1613
25
ML of clausal units (CUs) containing finite verbs Mono- lingual CUs % Biling- ual CUs % Total Cus % Welsh80784%19392%100085% English465%31%494% Either111%10%121% Dichot- omous 00%1 0%0%0%0%1 0%0%0%0% n/a97 10% 12121212 6%6%6%6% 109 9%9%9%9% Total961100%210100%1171100%
26
Summary of results on linguality and ML The vast majority of clausal units are Welsh The vast majority of clausal units are Welsh About 15% of clausal units are bilingual About 15% of clausal units are bilingual Clausal units with language-specific linguality mostly have the same language-specific ML Clausal units with language-specific linguality mostly have the same language-specific ML Clausal units with bilingual linguality mostly have Welsh ML Clausal units with bilingual linguality mostly have Welsh ML Dichotomous ML is extremely rare Dichotomous ML is extremely rare The model can be most exhaustively applied when the analysis is limited to clausal units with finite verbs The model can be most exhaustively applied when the analysis is limited to clausal units with finite verbs
27
Discussion of results Results of analysis show that : (a) MLF model is generally descriptively adequate in relation to our data (b) There may be a relation between the ML which is most frequent in a dataset and that which is favoured in bilingual clauses
28
Conclusions Use of clausal units, including both simple clauses and clause residues, makes it possible to identify linguality and ML of clausal units on a more principled and precise basis. Use of clausal units, including both simple clauses and clause residues, makes it possible to identify linguality and ML of clausal units on a more principled and precise basis. Use of clausal units in analyses of linguality and ML provide quantitative information which leads to hypotheses which can be tested on other data using identical methods of analysis. Use of clausal units in analyses of linguality and ML provide quantitative information which leads to hypotheses which can be tested on other data using identical methods of analysis.
29
Acknowledgements Arts and Humanities Research Council (AHRC), UK Marika Fusser Ellen Kimpton Research Councils UK (RCUK) Gary Smith
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.