As applied in the BNC, a lemma is a set of words consisting of a basic uninflected form (e.g. FLY) and its inflectional variants (in this case, flies, flew, flown and flying); these forms share the same part-of-speech category (in this case, verb). A 'Lemma query' is just like a standard query, except that it retrieves all the inflectional variants of a given word in one efficient search.

To find the lemma fly used as a verb, in the whole BNC:

1. Select Lemma query,
2. In the Lemma Query window enter fly in the first box, and under Lemma type select VERB. Then 'Start Query' (see below).

Other examples:

• To find fly used as a noun, in either singular and plural form, change the lemma type to SUBST (short for 'substantive').
• To find happy in its positive, comparative and superlative forms, i.e. happy, happier, happiest, simply type happy in the "Enter word" box, and choose ADJ (for adjective) in the Lemma type box.
• To find irregular, including suppletive, forms of nouns, verbs, adjectives and pronouns, enter the base form (i.e. the dictionary headword) as the lemma, as follows:

  Lemma word	Lemma type	finds
  GO	VERB	go, goes, went, gone, going
  GOOD	ADJ	good, better, best
  CACTUS	SUBST	cactus, cacti
  SHE	PRON	she, her, hers

Just as with standard queries, Lemma queries are applied to the whole corpus, unless you specifically ask for written texts or spoken texts only. You can do this in two different ways:

• In the basic Lemma Query box (as illustrated above), under Restriction, change 'None (Raw Input)' to 'Spoken texts' or 'Written texts'.
• Under Lemma queries in the left-hand menu bar, choose either the link marked Written texts or the one marked Spoken texts. You specify the lemma and its type in the usual way, and then select different sub-categories of spoken data, e.g. in spoken texts, choose 'Dialogue' as the interaction type. If no categories are chosen, the whole spoken or written BNC will be selected.

1. As with Standard queries, the concordance output sometimes shows cases where the grammatical tagging of the search expression is marked as ambiguous. For example, the query above, looking for FLY used as a verb, finds not only "clear" verb uses (i.e. clear according to the grammatical tagging program), but also lines with ambiguity tags: the first entry in the query result window has flying tagged with the ambiguity tag VVG-NN1:
   Keep the 30th Anniversary log <VVG-NN1>flying around the country for Amnesty's anniversary. (BNC A03.253)

2. Lemma queries do not find cognate forms belonging to different parts of speech, for instance, wonderful (adjective) and wonderfully (adverb). These forms can, however, be retrieved using a Standard query, as follows:
   (wonderful|wonderfully)

3. Unfortunately, lemma queries do not find variant spellings of words, such as the -ise and -ize suffix alternants of many verbs. Once again these can be found with a Standard query, e.g.
   (specialise|specialises|specialised|specialising|specialize|specializes|specialized|specializing)

   Alternatively use the "Lookup a word" function available from the top-level menu (under Other Functions). In the box next to Word or pattern enter speciali, and next to Entry type choose "beginning of word", and for Lemmatisation choose "Lancaster".