Quoting linguistic examples

In most linguistic papers it is both relevant and important to quote examples from a corpus or from other material you have studied. Following good academic practice, you should always state where the example comes from.

It is usually convenient to number the examples so that you can refer to them as e.g. "example 2" or simply (2) when you discuss it. This is better than talking about e.g. the following example or the above example in case you reorganize your paper or in case the reference is not entirely clear. There may for instance be more than one relevant "above example".

The below extract from Hasselgård (2017) shows how examples may be rendered (you may also see how this has been done in research papers and books that you read as part of your courses or your own research).

Note particularly the following:

• Both examples are numbered.
• Each example is followed by a reference saying where it was taken from, in this case the VESPA corpus. The abbreviation LIT says that the examples are from the literary analysis part of the corpus. Sometimes it is relevant to give an even more precise reference to the exact place in the corpus the example is found (for example a unique text code and even a sentence number).
• The numbers preceding the examples may be enclosed in brackets, as shown above, but they need not be. It is, however, common practice to use brackets when example numbers are referred to in running text. For example, in Hasselgård (2017) the examples were introduced as follows: Examples (17)-(19) all signal that the reader is expected to share the writer’s opinion or evaluation.
• Some material has been highlighted with italics in both examples. These are the constructions that this study particularly focused on, namely expressions of obviousness. If you use such highlighting, you should explain it the first time you do it, e.g. in a footnote.

 Multilingual examples

If your examples are from a multilingual source, such as the Oslo Multilingual Corpus, it should be made clear which part of the example is the original and which part is a translation. This can be done simply by keeping the identification tags that are attached to OMC examples, as in the following (note that the T in the tag indicates that the example is a translation). An alternative system may be to always give the original first and the translation after.

1. Nor did he enjoy his meetings with Dr Forestier… (BC1)
   Han likte heller ikke konsultasjonene hos doktor Forestier … (BC1T)
   Lit: "He liked also not the consultations with doctor Forestier"

You may wish to include examples in a language that you do not expect all of your readers to know. In such cases it is useful to supply a word-for-word translation, also called a literal translation. Literal translations should be marked as such (note the "Lit." in example (1)), or it should be explained in a footnote where the translation came from; i.e. that it is your own!

In example (1), only a literal translation was supplied. If it had been important to show the Norwegian syntax more exactly, a more detailed version might have been chosen, for example writing "consultations.the" instead of "the consultations" to show that the Norwegian definiteness marker is a suffix rather than a determiner. 

Yet another solution is to provide an inter-linear word-for-word rendering (often called a gloss), which may include the explicit marking of morphology. The below example is quoted from Amfo (2007), who describes Akan, a language spoken in Ghana. Details of how to write glosses can be found here.

References

Hasselgård, Hilde. 2017   Stating the obvious: signals of shared knowledge in Norwegian-produced academic English. In Pieter de Haan, Rina de Vries and Sanne van Vuuren (eds), Language, Learners and Levels: Progression and Variation. Louvain: Presses universitaires de Louvain, 23-44.

(c) Hilde Hasselgård, ILOS, June 2020.