A short phrase may seem recognisable, but a familiar word or typo does not establish its author. The scope for authorship examination depends on how much independently produced written language the text contains, the circumstances in which it appeared and what it can validly be compared with.
Why message length matters
A single remark provides little material for observing stable linguistic habits. A formulaic greeting, forwarded quotation and one-word reply provide different information from an independently written account. There is no universal number of lines beyond which authorship is proven: the suitability of the particular material is assessed first.
Several messages are sometimes more useful than one, but they must not be combined merely to increase the volume. It is necessary to establish that they belong to the correspondence being examined, and common authorship must not be treated as already established.
What to preserve together with the text
- The complete message, with its original spelling, punctuation and paragraphs.
- Preceding and following remarks needed to understand the situation.
- Information about the platform, date, editing and method of obtaining the copy.
- Forwarded messages, quotations and suspected borrowed passages identified separately.
- The source export, if lawfully available, and an unaltered backup copy.
Which samples actually help
Comparison requires texts of known origin. Where possible, select independently written messages from a similar period, language and communicative situation. The same person's business statement and informal correspondence may differ substantially. A document edited by a third party must not be presented as a freely produced sample of that person's language.
It is useful to attach a brief explanation to each sample: who created it, when and under what circumstances, where the file came from and whether there are grounds to doubt that the text was independently produced. A sample inventory makes it possible to discuss comparability before the final question is formulated.
An example of the boundary between observation and finding
The same rare word occurs in a disputed remark and two samples. This is a reason to study the context of its use, rather than ready-made proof. The word may be shared professional vocabulary, a quotation or a reply to a previous message. The combination of features and possible explanations for the match are examined.
What to ask the specialist before work begins
Discuss whether the text is suitable for an identification task, which samples are missing and whether the messages can be examined separately. Content analysis, identification of the text's author and identification of an account owner are different tasks. A match in linguistic characteristics does not itself confirm who physically pressed the keys.
If the material is insufficient, a substantiated statement of the limitation is more useful than a promise of a categorical answer. Where necessary, linguistic analysis is coordinated with examination of the files' digital origin, without substituting one field for the other.