Dorian's study of East Sutherland fisherfolk Gaelic extends from 1963 to 2008. The study's central thread is that Gaelic variation does not support classic findings from the sociolinguistic literature that suggest that linguistic variation should be limited in very small and homogeneous communities, that this variation should be correlated with social stratification, and that any phonetic differences should influence a speaker's awareness of oppositions between linguistic variables.
The most striking phenomenon observed by Dorian is a high degree of indexically neutral variation, both inter- and intra-speaker, in the context of a socioeconomically undifferentiated, highly homogeneous population. What Dorian shows is that such a linguistic phenomenon cannot be understood without taking into account the sociological characteristics of fisherfolks' life in the area: families of fishers were poor, residentially isolated, and Gaelic-monolingual until the end of World War I. Since members depend on each other in a fishery, fisherfolk communities manifest kinship and egalitarianism, while the separateness of the community towards nonfisherfolk English-speakers is accentuated by endogamy and the fact that families constitute a workforce unit.
The linguistic variation Dorian observes is multidimensional: geography-related, age-related, style-related, and so on. Nevertheless, she observes the unwareness of alternative forms among speakers. A striking instance is provided by a sibling pair, with little age difference, who do not notice that in comparable sociolinguistic contexts they use blatantly different forms. An exception to that unawareness is related to the use of some kinds of items such as gratuitous English loanwords, which are negatively evaluated by the Fisher-Gaelic-speakers. For Dorian, this partial awareness shows that linguistic accommodation in the community involves not inter- or intra-speaker variation in Gaelic but rather language choice: speaking Gaelic establishes one's membership in the fisherfolk community (vs. nonfisherfolk). As Gaelic speakers are linked by strong kinship ties and social roles, they have no need to produce accommodation by adopting the interlocutor's linguistic forms, so the choice of variants remains free.
Dorian's claims are based on an analysis of eighteen linguistic forms in both elicited and freely spoken data, each of which has from two to five variants. She explores their correlation with age and gender and the percentage of each variant's use for each speaker. The primary finding is that variation remains what Dorian calls personally patterned. The pervasive presence of personally-patterned variation is a specific feature of the Gaelic-speaking fisherfolk community. It is transmitted to children through a social structure that always exposes them from the start of their social life to a large number of kin-group caretakers, in other words, to a large sample of linguistic variation.
Dorian's study offers an important contribution to the understanding of linguistic variation in socioeconomically unstratified societies where language is unstandardized. It highlights the difficulties of Western dictionaries in acknowledging the validity of co-existing forms, the relevance of the individual dimension for describing the linguistic system, and the implications of heterogeneity vs. invariance in language change or its absence.
