Flashcards
The Data
Cards in this deck (40)
- 1Data Collection depends on ...Our goals will guide us in the selection process which will lead us to choose different procedures.
(goal and research question) - 2TranscriptionImportant aspect of data collection involving spoken discourse
- 3TranscriptSpoken discourse turned into a document
- 4Aim of transcriptionObjectivity showing the different variables that intervene in the discourse studied
- 5Neutral transcriptionThere is not a totally neutral transcription, but the researcher should try to use a method of transcription that shows all the variables involved in the studied phenomenon
(words, sequential organization of the utterances, interruptions, pauses ...) - 6An important requirement to achieve a neutral transcriptionWorking with a sample of a "naturally occurring talk"
- 7Naturally occuring talkInformal conversation which would have occurred even if it was not being observed or recorded and which was unaffected by the presence of an observer or a recording equipment.
- 8ParsingAssigning labelled brackets to each constituent of a sentence
- 9Notation systemA set of convetions used to transcribe a spoken interaction
- 10Ideal notation systemThere is not a single accepted way and each analyst chooses the features that best suit the purpose of their research.
Any type of annotation presupposes a typology but linguistics lacks in agreement about the categories to be used.
The ideal procedure would be starting with a set of categories and use our own only in the cases that there is no agreement whatsoever - 11HIghly detailed transcriptionsHard to read and too distracting (too much information)
- 12Less detailed transcriptionEasy to read but not enough information
- 13Edward: 3 main principles to describe a transcription systemCategories
Transcripts
Mark-up for computational tractability - 14According to Edward categories must be ...systematically discriminable
Exhaustive
Systematically contrastive - 15According to Edward transcripts must be ...Readable to the researcher
- 16According to Edward mark-ups must be ...systematic and predictable for computational tractability
- 17Mark-upText-encoding system consisting of a set of symbols
- 18Ethical obligation of the researcherAs the researcher has more power than the other participants, he has the obligation to protect them, not to harm then in any way and to observe their legal rights
- 19CorpusA collection of linguistic data (written texts or transcriptions of speeches)
- 20Linguistic corpora are used as ....a starting-point of linguistic description or as a means of verifying hyothesis about a language
- 21Features of corpus-based analysis according to Bieber et al.Empirical (analysis of actual patterns of use in natural textes)
Use of corpora (basis for analysis)
Use of computer for analysis
Use of both quantitative and qualitative analytical techniques - 22Advantages of using corporaMore reliable
(using larger and more varied texts leads to conclusions based on frequency of use of a given linguistic pattern and it allows comparative analysis about usage in different varieties)
More objectivity (accurate descriptions of usage)
Inehaustable source of hypothesis about the way language works - 23Computerized corpora allow ...The storage and analysis of a much greater number of natural language texts that would be impossible if we had to store and analyze them by hand
- 245 Computer CorporaLondon Lund Corpus
Brown University Corpus of American English
Lancaster-Oslo/Bergen (LOB) Corpus of British English
The British National Corpus (BNC)
The International Corpus of English (ICE)
The Bank of English - 252 Online Computer CorporaThe Experimental BNC English
Shakespeare Online Corpus - 26Concordance Programs (definition)Programas that turn electronic texts into databases which can be searched
- 274 Concordance ProgramsWord Cruncher
TACT
SARA
WorSmith tools - 28Concordance Programs allow ....Searching for words combination within a specific range of words
Looking up parts of words
Accessing frequency lists - 29London Lund CorpusIt began in 1960s as The Project of English Usage and consisted of 200 texts of spoken and written material of 500 words each.
It has been now computerized - 30Brown University Corpus of American English1st computerized corpus in the history of linguistics (1960s)
Its aim is to represent a wide range of genres of published written texts in American English produced during a single year - 31Lancaste-Oslo/Bergen (LOB) Corpus of British EnglishCompiled in 1970s to match the Brown corpus using British English texts
- 32Reich's taxonomy of corpora according to 8 criteriaMedium
National variety
Historical variation
Geographical variation
Age
Genre
Open-endedness
Availability - 333 types of corpora according to mediumSpoken (London Lund)
Written (LOB)
MIxed (BNC) - 343 types of corpora according to national varietyBritish (LOB)
American (COCA)
International (ICE) - 353 typeds of corpora according to historical variationDiachronic (Helsinki corpus)
Synchronic (Brown, LOB and BNC)
One stage of language history (Shakespeare corpora) - 362 types of corpora according to geographical variationDialect sample (Scots)
MIxed (BNC) - 372 types of corpora according to ageAdult English
Child English (Childes) - 384 types of corpora according to genreLiterary texts
Technical English
Non-fiction
Mixed - 392 types of corpora according to open-endednessClosed/Unalterable (LOB)
Monitor (Bank of English) - 405 typed of corpora according to availabilityCommercial
Non-commercial
Online
On Floppy-disks
On CD-ROMs
Information
- Description
- B Discourse Analysis (2 The Data) Flashcards on The Data, created by Gabriella Foresti on 08/07/2023.
- Created by
- Gabriella Foresti
- Views
- 2
- Rating
More on Corpus types and analysis