7.
7 Practice and self-assessment
New methodologies to
investigate language change
© Cambridge University Press 2019
Quantitative versus
qualitative methods
• Quantitative (statistical) methods have been used in language
study for a long time to classify and count features and work
out percentages and frequency. Researchers gather data that
lends itself to statistical analysis.
• Qualitative methods have been used to identify and describe
features of language from real-life occurrences and examples
– developed from the ethnographic tradition of linguistic
research where researchers are participant observers.
Researchers gather data that lends itself to content analysis.
© Cambridge University Press 2019
New quantitative methods
• Computers and specialised software now play a
growing role in linguistics. Corpus linguistics focuses
on the repetitive patterns that can be attested in
corpora – large amounts of digitised data that have
been collected.
• Corpus linguistics enables huge amounts of data to
be processed very quickly and observations made
about things such as key words, semantic fields and
collocations (the tendency of particular words to co-
occur with others).
© Cambridge University Press 2019
N-grams
• N-gram – ‘n’ stands for the number of words that can
be see to occur in more than just random frequency.
• N-gram graphs are very useful to see how a word has
changed over time in terms of meaning/use.
© Cambridge University Press 2019
Concordance
• A concordance line brings together and displays
instances of use of a particular word from the widely
disparate contexts from which it occurs.
• A collocation consists of the core word combined
with words on either side. This means that patterns of
similarity can be detected in the words surrounding
the core word.
© Cambridge University Press 2019
New technology and patterns
• New technologies are particularly good for lexical
study and picking up patterns.
• Patterns that could not be easily picked up otherwise
can be uncovered.
© Cambridge University Press 2019
Concluding thought
Over the last fifty or so years corpus-based methods have
developed into one of the most rapidly growing and most
widespread ‘new’ technology in linguistics. Instead of
relying on intuitions of what can and cannot be said,
linguists are now turning more and more to corpus data
to see what is and what is not said. However, as is only
appropriate for a lively scientific discipline, it is still
evolving…
Stefan Th. Gries, 2014
© Cambridge University Press 2019