Tag: linguistics

Probably to my detriment, I like to join communities that discuss linguistics on social media. I don’t mean following others who conduct linguistic research — I do that, too — but rather communities where linguistics is discussed by anyone who happens to be interested in discussing it. I figure, “What’s the use of all the stuff I learn if I don’t share that with others, including the general public?” One online space that I peruse for this is reddit, and one topic that comes up there from time to time is that of creoles.

This is great, because I love talking about creoles and know a lot about them. However, it’s also not great, because I find that what I have to say is often disregarded or even unwanted (i.e., I get downvoted a ton). When creoles come up, there is almost always a misunderstanding about what they are that’s implicit in the questions people ask, and it seems that people don’t like it when I clarify what creoles are. What’s more, the misunderstanding goes beyond more casual forums such as reddit and even finds its way onto tightly run Q&A sites like Linguistics Stack Exchange (such as here and here). They do a bit better in the latter community, but still rely on the dated idea that creoles develop out of pidgins. Even among those who have taught at least an introductory linguistics course, it appears to be common that this is how creoles are conceptualized (Bancu et al. 2024). So, I thought a post here would be a good place to clarify what creoles are for those who are curious out there.

To start, let’s talk about what creoles aren’t. First, creoles do not develop from pidgins. This was indeed a common theory in the not terribly distant past, enough so that it found expression even in the writings on pidgins on creoles carried out be pre-eminent linguists from up to at least the 1980s (e.g., Halliday 1964/1968:143; Romaine 1988:38; Todd 1974/1990:2). It was so embedded in the minds of creolists at that time that Thomason & Kaufman (1988:148) and Arends (1994:15) still felt the need to add it as a type of creole even while acknowledging the sociohistorical definition of creoles that makes far more sense. There have been a number of solid arguments made against the pidgin-first theory (Mufwene 2001:7-11), a major one being the wide admission that there is no textual evidence of a pidgin stage for any known creoles (Aboh & DeGraff 2017:413). Additionally, this theory also obscures the overall picture. To accept that a pidgin must develop first and then gain native speakers raises all sorts of questions about what it means to be a native speaker, how many native speakers count, etc. Perhaps the biggest obscuring caused by the theory is that pidgins have traditionally been defined as coming out of the context of trade and creoles, as we’ll see, have not been defined that way, so the sociohistorical picture is muddied by combining these two categories.

The pidgin-first theory is what those with some knowledge but who are not experts often go to. Another faulty approach, though, is to define creoles structurally, which tends to be the preferred method of laypeople and a very select few linguists. By select few, I mean almost exclusively McWhorter (1998) who argued that there were three linguistic structures that justify classifying creoles as a distinct type of language: no inflectional affixation, no tone, and semantically regular derivation affixation. The problem is that he himself pointed out that even the creoles that he chose as his prototypes had exceptions to some of these structures, leading him to implicitly argue that they simply need to be close enough to meeting these structural descriptions, but then what is “close enough”? How do we quantify that in any way that isn’t arbitrary? It’s a fine needle to thread, and unsurprisingly this approach has been heavily critiqued by other creolists.

The definition of creoles that experts actually use is that a language can be classified as a creole if it was born out of a specific sociohistorical context, namely that of the subjugation and exploitation of one people by another (i.e., slavery or slavery-like conditions). Because of the popularity of pidgin-related and structure-related definitions of creoles, this may appear to be a new theoretical approach, but it’s really not. Perhaps the earliest scholar who could be called a creolist embedded this sociohistorical context into how he defined creoles (Reinecke 1938). This focus on the sociohistorical continued from there (Ansaldo & Matthews 2007:8; Baker 1990, 1994; Bartens 2013:65; Meyerhoff 2018:279; Mintz 1971:481; Winford 2003:308). Even for someone like Bickerton (1988), who is infamously attached to structural arguments about creoles, still relied on sociohistorical context to classify creoles.

It is quite difficult, actually, to find a creolist whose definition of creoles does not refer to sociohistorical context. There was a tendency in the early 20th century to speak euphemistically about slavery, but the point is always clear once you see it. For those who do mention slavery explicitly, it still can seem uncomfortable for them at times, either on social or theoretical grounds. Hence, you also find scholars hedging their definitions a lot, stating that creoles “usually” implicate slavery or something along those lines. However, examples of creoles that don’t implicate slavery are not then produced. For instance, Thomason & Kaufman (1988) highlighted slavery but did not believe it was present in all cases where a creole developed. They actually did give an example, that of Pitcairnese (148). This creole came out of a mutiny in 1790 on an English ship that had enlisted the help of Tahitians. After the mutiny, the crew settled on Pitcairn Island to avoid repercussions, after which the creole developed as the English crew and the Tahitians learned to communicate with each other. Thomason & Kaufman (1988) described their habitation as “egalitarian”, but they failed to note that the Tahitians were treated like property to such an extent that there were eventually murders committed in their small society in retaliation, painting a picture much closer to slavery and thus not the counterexample that Pitcairnese was supposed to be.

So why does this matter? It matters because the way we define our object of study leads us to ask very different questions and can prevent us from elucidating answers that we’re looking for. Defining creoles as coming out of the sociohistorical context of slavery leads us to ask sociolinguistic questions about power, race, colonialism, etc. Definining creoles as coming out of pidgins leads us to spending a lot of time on trying to answer questions about how a pidgin becomes a creole, something that we can’t answer if it doesn’t happen, or simply obscures our abililty to understand creoles or pidgins. Even worse, defining creoles in terms of structure is a form of exoticizing these languages and, by extension, the people who speak them, people who are often stigmatized to begin with. This exoticizing or exceptionalizing is something that DeGraff (2003, 2004) has spoken out against for quite some time but, unfortunately, has been slow to filter through to all relevant parties.

Thomason, S. G., & Kaufman, T. (1988). Language Contact, Creolization, and Genetic Linguistics. University of California Press.

Todd, L. (1974/1990). Pidgins and Creoles (2nd ed.). Routledge.