Observer's Paradox

The Observer's Paradox is a theory proposed by William Labov, the father of variationist sociolinguistics. It refers to the difficulty of extracting natural speech from informants in order to analyse contemporary patterns of use. *If a fieldworker requests an interview with a native speaker and proceeds to ask questions and record the responses, the informant, aware that everything they say is to be analysed by a linguist, is likely to adapt their speech, make it more "correct", use a higher register and so on. This inevitably leads to hypercorrection in the realms of phonology and the shying away from the vernacular which is precisely what variationists are aiming to study. The paradox, therefore, lies in the fact that there needs to be a linguist with a microphone to capture speech but the presence of said linguist with microphone will necessarily provoke the 'wrong' type of speech.

Let's look at this in a little more detail

First of all, let us take a short detour in order to distinguish between the native speaker and the naïve informant. The latter is used mainly by syntacticians. The informant, a native speaker, is presented with a list of sentences and asked to give judgements on each sentence's acceptability. In other words, whether he considers the sentence to be grammatically correct or not, and perhaps if the informant would use the same expression and if not what would he say instead. The sociolinguistic interview also relies on native speakers but language is not discussed with the informant, given that the goal is to obtain natural contemporary speech. A conversation takes place and is recorded for later analysis by the linguist, according to his or her preferred methods. So, for syntacticians, the Observer's Paradox is unimportant while for sociolinguists it is a crucial hurdle to be overcome.

The 'vernacular' is the term used in variationist sociolinguistics to describe the natural spoken language of the target community. In other words, how people actually speak when in their natural environment. So not your telephone voice, or how you speak at a job interview, but how you speak with your peers. Though inevitably intertwined with regional dialect or accent, the vernacular has the same definition whether you are from South Carolina, Perth or Liverpool. Further discussion goes rather beyond the topic of this write-up, suffice to say that within a regional dialect there exists a vernacular and at the same time there are elements of vernacular speech which cross continents and languages.

Hypercorrection occurs where a speaker is aware of certain grammatical or phonological rules but their vernacular does not use the prescribed rules. At times the need to speak 'properly', in a formal situation perhaps, means that our vernacular speaker has to consciously delve into his syntax memory as the rule is not part of his hardwired grammar (as yet, there appears to be no node on Universal Grammar, so in the meantime, see Gritchka's write-up on Chomsky for more detail on this). The usual outcome is that he will apply the rule rather overzealously, which often results in incorrect use. As an example of this, let's take a speaker of Cockney, a dialect known for h-dropping (the non-pronunciation of the 'h' at the beginning of a word) as in A for 'orses, rendered as Hay for Horses in Standard British English. Speakers of Cockney know that SBE does not drop the 'h' and when trying to adjust their dialect towards the accepted standard, a process known as accommodation, they would call up this knowledge and sound their aitches. So, by extending the rule to all words beginning with 'h' they end up pronouncing words such 'hours' with an audible 'h'. In reality, speakers of standard English do not pronounce the 'h' in 'hours' and so our Cockney speaker is said to hypercorrect, that is overextend the rules.

How to sidestep the Observer's Paradox

How Labov chose to avoid the impact of a professional linguist sitting across the table from his interviewee was actually quite simple. Whenever possible, the fieldworker should be of in-group status. In other words a friend or acquaintance of the informant. This automatically alleviates some of the awkwardness and promotes a more natural environment. Next there is the favoured option of interviewing self-selected dyads. So, perhaps a couple, a small group of friends or a family. Such a group is unlikely to alter their speech excessively, given that they have pre-established ways of talking to each other. The third card Labov had up his sleeve was to ask the informants to recount tales of personal experience. He would ask questions such as Were you ever in a situation where you were in serious danger of being killed? Any story which provokes deep emotional responses from the teller will, he posits, produce natural speech. If the informant is simply going through a list of what he did that day then they are free to concentrate on their speech. However, if they are overtaken by memories of an extremely poignant time in their life, there is much less likely to be a checking of the language (be that grammar, vocabulary or accent). And in this way, it is possible to circumvent the Observer's Paradox and to achieve the objective, as Labov said, of "tapping the vernacular."

brought to you in association with: a social sciences quest, nodeshell rescue, node your homework and linguistics 101 with princess loulou.

*ariels pointed out that the reference to the Heisenberg Uncertainty Principle is misleading, so I have removed it.

References: