The excessive variation in definitions means they cannot be differentiated

The argument

There are several definitions that have attempted to capture the distinction between a language and a dialect, but there are too many linguistic varieties that do not neatly fit into these definitions for them to be valid. One criterion that supposedly defines two or more linguistic varieties as languages and not dialects is that they are not mutually intelligible--that is, the speakers of one language should not be able to understand the speakers of another without training. By this definition, two people who speak different dialects of the same language should be able to understand one another without any issue, but this is not always the case. Norwegian, Danish, and Swedish are all widely considered to be separate languages, yet all three are mutually intelligible. And Mandarin and Cantonese are often classified as dialects of Chinese, but the speakers of one cannot understand the other. The other main criterion used to distinguish languages and dialects is best expressed in a quote by sociolinguist Max Weinreich: "A language is a dialect with an army and a navy." Essentially, this means that languages are official, standard, written, and prestigious while dialects are not. However, there are several problems with this viewpoint. This definition perpetuates the belief that the standard variety (language) is inherently better and that vernacular speech (dialect) is less sophisticated. Also, many nonstandard dialects are written down, both in linguistic research and even on social media platforms like Twitter. Since none of the previously established criteria can differentiate between languages and dialects, we can only say that language does not exist--only dialects.

Premises

Counter-arguments

This view erases the fact that both the linguistic and sociopolitical definitions can account for much of the differences between languages and dialects although more research in this area is needed to fully distinguish the two. Also, saying that only dialects exist and not languages is unhelpful in many ways. For one, it would be difficult to analyze and discuss language families and how different languages are related, especially for a language like English that has many dialects. It is much easier and more efficient to unite all a language's dialects for the purposes of analyzing large-scale language change through time, for example. This "solution" would also not necessarily change the fact that some dialects have more social power and prestige than others, and many dialects would likely remain stigmatized. Finally, current research suggests that human brains process bilingual code-switching differently than dialectal code-switching. Since the mind recognizes language as a valid phenomenon separate from dialect, linguistic theory should do the same.

Rejecting the premises

[Rejecting P3] The move from imperfect criteria to no distinction is too strong: the counter holds that the linguistic and sociopolitical definitions between them account for much of the difference even if neither is exhaustive, so the appropriate conclusion is that more research is needed rather than that one of the categories is empty. [Rejecting C] Abandoning 'language' also fails on its own terms — it would make language families and historical change difficult to describe, particularly for a language with as many dialects as English; it would not remove the prestige asymmetries the argument objects to, since stigmatised dialects would remain stigmatised; and the counter cites evidence that the brain processes bilingual code-switching differently from dialectal code-switching, which treats the distinction as a real phenomenon rather than an artefact of definition.