Alan Reed Libert

Grammatical Viruses and Real Ones: Implications for Endangered Languages

Alan Reed Libert

University of Newcastle, Callaghan, Australia


Abstract: This paper looks at a grammatical phenomenon which has been referred to as a grammatical virus. An example of this in English is the nominative case form of pronouns in constructions such as between you and I. It is argued that real, i.e. biological, viruses, such as the one responsible for COVID-19, may create situations from which both grammatical viruses and endangered languages (as well as other minority and heritage languages) may benefit.

Keywords: COVID-19, grammatical viruses, endangered languages

1. Introduction

The world has been dominated by the COVID-19 situation, and the negative effects on humanity have been huge – many lives lost, many people sick, many jobs lost. Without minimizing these negative effects, there have also been some positive effects; for example, air pollution has (for a relatively short time) diminished in some places. Another possible positive effect involves endangered languages (and other minority languages and heritage languages): many people have been in lockdown, i.e. at home, and the home is where many such languages are primarily (or exclusively) used; children in homes where a minority or heritage language is spoken might have more exposure to that language than they would normally (e.g. when they are at school).

2. Grammatical viruses

COVID-19 is caused by a (quasi-)biological entity, a virus. I will now talk about a quite different kind of virus, a grammatical virus. The notion of a grammatical virus was introduced by Nicholas Sobin; below is a quotation from him about these viruses.

“I propose a theory of editing or monitoring toward prestige forms that I call virus theory. The particular rules that facilitate such editing toward prestige constructions are called grammatical viruses... A grammatical virus is a device that can read grammatical structure and affect it, though it is grammar-external. A virus is parasitic on a grammar, and, as I will argue, it facilitates the construction of prestige forms.” (Sobin, 1997, p. 319)

Later on the same page he gives a possible example of a construction caused by such a virus:

(1) ... between you and I... (Sobin, 1997, p. 319)

Of course, the expected or prescriptively correct form of this would be … between you and me …, since in English prepositions (usually) take accusative forms of pronouns. However, examples such as (1) are not very rare. As Sobin (1997, p. 319) indicates, such constructions are due to hypercorrection. One can easily imagine how they would arise: when children say things like “Me and John are going to the park” or “John and me are going to the park”, they are told by their parents to say “John and I are going to the park”. The children would then posit a rule requiring the form I rather than me after the word and (regardless of the larger context, e.g. whether the conjoined noun phrase was a subject or a prepositional object). This rule would then remain in the children’s linguistic system as they become adults.

This reference to and is important, as it displays a key feature of grammatical viruses; Sobin (1997, p. 329) states, “The first property that the prestige [i.e. virus] rules exhibit is lexical specificity: a virus strongly involves particular lexical items. Lexical specificity is uncharacteristic of mainstream syntactic processes, which are normally lexicon-neutral and category-neutral”.

Some other examples of sentences created by viruses are given in (2); they come from Sobin (1997, p. 318), who says, “Nonprestige forms [i.e. forms not due to a virus] are given in parentheses”.

(2a) Mary is richer than I (me).

(2b) There are (-’s) books on the table.

(2c) There are (-’s/is) a cat and a dog in the yard.

Lasnik and Sobin (2000) see whom in English as involving a virus; I will return to this word below.

Consider now the following passage:

“Though viruses are external to the grammar, they are nonetheless ‘natural’, and can be ‘smoothed in’, so that the output may sound natural or acceptable to a native speaker. However, unlike grammatical principles, viruses must be explicitly learned because, in their natural form, they are lexically specific. Viruses are, in effect, lexically concrete syntactic idioms. Unlike the normal principles of grammar, viruses are acquired and applied in connection with particular vocabulary.” (Sobin, 1997, p. 338)

To me this sounds a little like the acquisition of a second, i.e. non-native, language. The passage below expresses the same sort of idea:

“The grammar of a standardized language is thus heterogeneous, in that it comprises both generalized rules acquired through normal acquisition and highly specific pseudo-rules that are picked up at school or through contact with speakers who themselves have learned the relevant pseudo-rule. Virus Theory … captures this duality by positing the existence of parasitic rules or ‘viruses’ that colonize the basic grammar and manipulate its output.” (MacKenzie, 2013, p. 21)

2. Grammatical viruses, endangered languages, and biological viruses

I would suggest that there may be similarities between virus rules and endangered/minority/heritage languages, in the sense that both may be acquired “artificially”, i.e. not through the normal first language acquisition process. Both may require some effort: it might take effort to acquire an endangered language, if it is not one’s first language, and it might require some effort to use it (especially outside the home), even if it is one’s first language, when the main language in the community is another language, i.e. one might have to remember to use the minority language. In the early stages of acquiring a grammatical virus rule, a (probably young) speaker might forget to apply it – it might take several corrections by an adult before it takes hold, and it might take some effort on the part of the speaker to remember it.

Especially given the current world situation, I should hasten to add that, when I link grammatical viruses and minority languages, I am not at all implying anything negative about minority languages. It should perhaps be pointed out that not all biological viruses are bad (although the most famous ones certainly are); there are viruses which are good (from a human point of view), e.g. bacteriophages.

However, there is a more significant possible similarity between minority/heritage languages and grammatical viruses: perhaps both might be affected, either adversely or positively, by the lockdown situations which we have seen and are seeing in many parts of the world. Consider the situation with whom. I think that I am in a small minority of English speakers, but I try to use whom in the appropriate contexts, e.g. in “Whom did you see?” (Although I am a linguist, and linguists are supposed to be descriptive, not prescriptive, I am extremely prescriptive.) Not only do I use it myself, but I might correct other family members when they say who in such contexts.

I am probably the only person whom [!] my children are in contact with who does this (I would think that their teachers at school do not do it), and so they do not get that much exposure to it, although of course I spend a lot of time with them. However, during the school closures due to COVID-19, I spent more time with them than I was before, and there was thus the possibility that they would experience my corrections with respect to whom more often. There probably were not more corrections during this time, since it is not a correction which I necessarily make every day or even every week, but there was that possibility, as well as the possibility for other corrections towards prestige forms. In that sense, the situation caused by COVID-19 may have been beneficial for the grammatical viruses of English, giving them more possibilities to take hold, one might say.

Turning now to endangered, minority, and heritage languages, the situation may be similar in at least some households. If children are not going to school, or many other places, due to lockdown requirements, they will be spending more time at home, and the home is where such languages will probably be spoken most. At least for a brief time they may not be so exposed to majority or dominant languages, which will probably be to the benefit of their home (i.e. minority or heritage language).

4. Conclusion

For more than a year, the world has been struggling with the COVID-19 situation. We have faced enormous challenges and tragedies, but there may also have (relatively minor benefits) for endangered, minority, and heritage languages: just as the biological virus may have assisted in the survival of some grammatical viruses, it may also have played a small role in the preservation of these languages.


Lasnik, Howard and Nicholas Sobin (2000). The who/whom puzzle: On the preservation of an archaic feature. Natural Language & Linguistic Theory, 18(2), pp. 343–371.

MacKenzie, Ian (2013). Participle-object agreement in French and the theory of grammatical viruses. Journal of Romance Studies, 13(1), pp. 19–33.

Sobin, Nicholas (1997). Agreement, default rules, and grammatical viruses. Linguistic Inquiry, 28(2), pp. 318–343.