POS (or ‘part of speech’)

POS (or “part of speech”) is used in Sketch Engine in two senses:

  1. It is used generally to refer noun, verb, adjective etc.
  2. POS can also refer to the POS attribute which is explained below.

Tokens in a corpus are assigned various positional attributes. One of them is POS. It contains the part of speech only and does not include any additional information. (Do not mistake it for tag.)  POS is the one-letter lempos suffix. POS may not be present in all corpora. You can check the POSs (lempos suffixes) on the corpus info page of each corpus.

Here are some examples of attributes:

word lemma tag POS lempos
dogs dog NNS n dog-n
dog dog NN n dog-n
waiting wait VBG v wait-v
longest long JJS j long-j
this this DT x this-x

Specific lempos suffixes, and therefore POSs, only exist for main categories. All remaining categories are assigned a special POS “x”.

POS is especially useful in a situation like this:

Task: Is reply more frequent as a noun or verb?

The tool to use Concordance – Frequency. The frequency can be counted using the tag attribute or the POS attribute. This is the result when using the tag attribute –  verbs are divided into present/past/gerund/third person etc. and nouns are divided into plural/singular. This is impractical.

tag Frequency
NN 1,328
VVD 933
VV 394
NNS 205
VVZ 150
NP 125
VVG 85
VVN 53
VVP 40
NPS 1
To get the answer, the user would need to manually add up all the nouns tags and verb tags. Instead, the Frequency should be counted using the POS attribute to get this:
pos Frequency
noun (n) 1,659
verb (v) 1,655

This is more compact and more practical. The user gets the answer at first glance.

Compare

tag

See also

lempos

positional attributes

« Back to Glossary Index