Intro to Corpus Linguistics

· @tasali · ELIN479, Week 3-1

Definitions of a descriptive grammar, Unit 2.

How we understand grammar seems to be related to how we construct it. Grammar by default is not prescriptive as what we mean by rules changes and is often broken by exceptions, which all natural languages experience. The spoken grammar is also different from that of written as one of the core values for the spoken grammar is efficiency.

Descriptive grammar looks for an answer to what is probable. Our view of grammar is descriptive rather than prescriptive.


Discourse is a text contextually coherent with more than one sentence.

In spoken language analysis, an utterance is the smallest unit of speech. It is a continuous piece of speech beginning and ending with a clear pause. For, instance, “Stop!”

Grammar is the naturally evolved rules governing the language production in the case of a natural language. “We seek to base our ideas on evidence of actual language use, which we can find in corpora. This does not mean that anything goes but rather that we use corpora to help define such aspects of grammar as rules” (Jones & Waller, 2015, p. 20) 1.

Morpheme is the smallest meaningful unit in a language.

Content morphemes express a concrete meaning or content, and function morphemes have more of grammatical role. For example, the morphemes fast and sad can be considered content morphemes. On the other hand, the suffix -ed is a function morpheme since it has the grammatical function of indicating past tense (“Morpheme,” n.d., para. 10) 2.

Definitions of a descriptive language

Looking at corpora, we will see that strict rules do not always apply languages in use and there are always exceptions.

A sentence may consist of a single clause, multiple clauses joined by a conjunction, or relational clauses where a single clause has subordinate clauses expressing wider range of meaning.

I follow rivers when I am sleepwalking.

The sentence above has an independent (main) clause and a dependent (subordinate) clause. A dependent clause will not make sense on its own (missing context and main clause) as it is for supplying the main clause with information whereas a main clause makes sense with or without dependent clauses.

This will not make sense:

When I am sleepwalking.

This will make sense:

I follow rivers.

The grammar in a speech is different from that of that is written. This could also show how we speak and write in corpora. Together with utterances, a discourse marker will help construct sentences that make better sense in speech. “A discourse marker is a feature of a spoken language that is used to rephrase or clarify an utterance in conversation” (Jones & Waller, 2015, p. 28) 1. Let’s take the following dialogue as an example to this:

aYou are joining us today, right?
bFor sure!
bI mean, if my mother lets me.

In the second exchange, b says, “For sure!” which translates to “Yes, I will join you today if my mother lets me.” We can extract this meaning because the second exchange is an utterance that will only make sense in a context and it will address the previous exchange. In the last exchange, b expands from the utterance and “I mean” is a discourse marker clarifying the previous message.

The choice of words and how they are structured in a message set the tone of. To understand and analyze it, “Halliday and Matthiessen (2004, 2013) suggest that all texts can express ideational, interpersonal, and textual metafunctions. The ideational metafunction means that language is used to help us understand an express our ideas and perceptions of the world. The interpersonal metafunction suggests that it is used to enable us to participate in communicative acts with other people, to maintain, build and establish relationships, and and the textual language is used organizationally to relate what is said (or written) to the text itself and the real world outside the text” (Jones & Waller, 2015, p. 27) 1. Take the following text as an example:

aAre you even listening to me?
bHardly ever!

b says ideationally s/he is not listening to a and interpersonally, the b shows no enthusiasm and, thus, is rude.

Single words have their own grammar, e.g, my book or you book. In “my book” is ‘my’ a determiner and indicates possession, but in the case of “you book”, ‘you’ is not a determiner and does not indicate possession. “Words also cluster together in predictable ways to make units which have their own internal grammar. To recognize this, in this book we often use the term lexico-grammar after Halliday and Matthiessen (2004: 43). They suggest that lexico-grammar is a term we can use to refer patterns which lie somewhere in between structures and collocations, having some of the properties of both…. Sinclair (1991) termed this phenomenon the idiom principle and suggested that much language is learnt and used as prefabricated wholes, in the way we would learn and idiom such as “I’m over the moon.” This was in contrast to the open choice principle, by which we use grammar to generate unique utterances, sentences an clauses” (Jones & Waller, 2015, p. 28) 1. Most languages consist of these types of chunks which, instead of consisting unique choice of words, are relational and grammatically broken when analyzed but makes sense as a group in our generic grammar knowledge. To give an example, “Will you marry me?” is grammatically incorrect, but we prefer it over the grammatically correct “Will you be wedded to me?”. Also, we could also use “would” instead of “will”, but some words often co-occur with some other words. This is called collocation where a word is primed as the language production has come to be that way and not due to grammar not being used.

this unit should have covered until the page 34, but we studied until page 30 due to lack of time. Not sure if we will return to it later.

  1. Jones, C., Waller, D. (2015). Corpus Linguistics for Grammar: A Guide for Research. Routledge.  2 3 4

  2. Morpheme. (n.d.). In Wikipedia. Retrieved October 24, 2020, from