memory

remembrance of
seen smile
heard cries
come to nought
without input
into this vast machine

detach the lines
vanished functionality
love & hate –
a memory
with no outlet
for your emotion

earring

Also ear ring. A humming sound the middle-age hear (“here”) that seem to come from no particular direction (“here”).

KWIC

Short for Key Word In Context. It is a way of looking at a search term (type) in a concordance program with the keyword centred so as to see the patterns created by the other words, its context.

Below is an example of a concordance search of the term ‘violence’ in a corpus.

violence kwic

The words ‘domestic’, ‘TV’ and ‘of’ seem to stand out and warrants further investigation. This is even before the surrounding text has been sorted.

Other Corpus Linguistic Terms:

Five or Seven Sentence Patterns?

While the seven sentence pattern description is the norm in English linguistics today there still persists the use of five sentence description in some non-English speaking countries like Japan which teach English as a foreign language.

Essentially the seven sentence pattern is a five sentence pattern with the extra two pattern as extensions of SVA and SVOA. The problem is that some common sentence patterns seemingly cannot be described by the five sentence pattern model. Take sentence (1.), for example:

  1. John sat up.

There is ‘John’ and he is performing the action of sitting up from perhaps a slouched position. In other words there is one actor doing one action. Therefore it is an SV pattern (John (S) / sat up (V)). Now consider (2.):

Continue reading

type

The unique form of the tokens (words) in a corpus. Often accompanied by frequency data.

Meaning is treated as secondary. Corpus linguistic analysis does not directly reveal the various meanings of a word. This must be inferred from its usage. In corpus linguistics this usually done by concordancing, collocations, clusters, etc.

token

The individual forms (words) of a corpus. The sum of the tokens is the size of the corpus. The term contrasts with type in order to distinguish how we are observing the form, whether as one instance in the corpus (token), or as combined instances relating to its frequency within a corpus (type).