Dylan Scott Low Y.J.

dylanscottlow at outlook dot com

© 2024 Dylan Scott Low Y.J.

Hosted on GitHub Pages — Hydeaway theme by jpatrickdevine

​

08 March 2024

Preamble: The following consists of a summary/personal interpretation of my reading of Chapter 1 “Words and Meanings: The Need for a New Approach” from Patrick Hanks’ monograph Lexical Analysis: Norms and Exploitations (2013: The MIT Press).

Hanks, a lexicographer who worked on the Collins COBUILD dictionaries, presents an interesting corpus-informed theory of language that dispels with binary categories such as grammaticality and replaces them with continuous variables (categories of degree) which can be quantified with the help of native speaker corpora.

I develop an understanding of Hanks’ framework by thinking through a key problem in Singapore English research — how to conceptualise the fuzzy boundary between Singlish and standard English (something I was thinking about after a recent conversation with Nick Huang at the National University of Singapore).

I arrive at a tentative conceptualisation of this problem using Hanks’ terms at the end, though by no means is this conclusive. But I think it’s a good first step to build off of. I also end with some interesting quotes from Hanks’ chapter.

All grammatical sentences of standard English are also potential grammatical sentences of Singlish. That is, any well-formed standard English sentence (e.g., 1a) can also be judged to be a well-formed Singlish sentence. This is a problem, since it is then easier to distinguish sentences of Singlish (which may be exclusive to Singlish, e.g., 1b) from sentences of standard English but not the other way around — an asymmetry.

 1a. There is a house on the hill over there. (Standard English or Singlish)
 1b. Over there the hill got house. (Singlish only)

However, perhaps attempting to capture the set of all possible grammatical sentences (following Chomsky, 1957) is the wrong approach. As Hanks (2013, p. 7) sharply remarks, “Linguists who seek to explain all possibilities in a single step have failed, will fail, and are bound to fail.” The solution which Hanks advocates for, then, is to focus on probabilities rather than discrete categorisations of what is grammatical and what is not. This is one of the fundamental ideas underlying Hanks’ Theory of Norms and Exploitations (TNE).

Through TNE, Hanks offers a two-step process as an alternative. First, identify the normal, central, and typical uses from a native speaker corpus. This is an approach that follows from Rosch’s prototype theory in cognitive psychology (also prominent in Cognitive Linguistics), where some instances of a conceptual category are more prototypical than others. Hence, prototypicality (or simply, “typicality”) is a quantifiable property.

Secondly, we then see if non-typical usages which do not fit well-established patterns can be explained using a different set of rules (i.e., irregular rules). We should not confuse typicality of use with frequency of use (Hanks also calls this “social salience”). The difference between rules that apply to typical uses and those that apply to non-typical uses is just that: a matter of difference in typicality.

Hanks additionally argues that social salience (as frequency of use) and cognitive salience (as ease of recall) are distinct variables from one another. He also hazards a stronger hypothesis: that social salience and cognitive salience share an inverse relationship.

(Side note: This stronger hypothesis seems to be contradicted by my personal intuitions about Singapore English—in colloquial Singapore speech, standard English constructions are more socially salient, i.e., infrequently produced, but they do not seem more cognitively salient, at least in my subjective experience. Perhaps we should distinguish social salience in production and comprehension as distinct.)

For Hanks, language acquisition involves gaining two different but interconnected rule-governed competence systems. First, there is the acquisition of norms (i.e., typical usages, including idioms). Second, there is the acquisition of exploitation strategies, wherein norms may be exploited in various ways to achieve new meanings. Hanks (2013, p. 23) states: “The rules governing exploitations are different in kind from the rules governing conventional usage.”

Hanks uses the preposition like as an example of a semiotic resource that English speakers use to exploit norms. For example, in a sentence such as Seals are like dogs that live in the ocean, the simile marker like makes it known that the term dogs is being exploited beyond its normal (literal) range of referents. Here, like highlights certain properties of the meaning of dogs which are shared with Seals, though it also suspends others (most explicitly, that dogs do not live in the ocean) such that Seals are not taken to be normally under the category of animals we call dogs.

Other exploitation strategies include ellipsis, novel conceptual metaphor, or new lexical coinages using combining morpheme forms (e.g., Kafka + esque = Kafkaesque).

Hanks emphasises the role of collocations in determining whether a lexical item is being exploited or used normally. He cites Sinclair’s (1991) proposal of the open-choice principle and the idiom principle. The open-choice principle sees language as a result of many complex choices, whereby the choices are only constrained by the criterion of grammaticality. This view is similar to that of Hallidayan Systemic Functional Linguistics, which takes the modelling of such open choices (using system-network notation) as a fundamental task of linguistic description.

The idiom principle, in contrast, refers to the role of language-internal resources (constructions) which constrain particular choices in the language-generation process. Hanks demonstrates with this example: when the verb switch is used with some kind of mechanical device, it has a marked idiomatic preference for the particles on or off to follow it. Yet, in its other sense of ‘to replace one thing with another’, it is far more flexible in its collocations.

Applying Hanks’ TNE to the case of Singlish and its asymmetrical relationship with standard English, it seems appropriate to characterise Singlish grammar (at least, those norms which are socially salient from the perspective of non-Singapore standard English speakers) as a set of additional norms not available to speakers who only speak standard English. This explains why standard sentences are grammatical in Singlish but not the other way around; Singlish is English plus more.

Now, this does not mean that standard English constructions are more typical in Singlish. In many social contexts, especially when “basilectal” Singlish is in use, standard English constructions would be non-typical (thus, infrequent and socially salient). The same could be said for the speech of “monolingual speakers” of Singlish, that is, people who speak Singlish but lack proficiency in standard English. Such speakers learn the “plus more” part of Singlish (i.e., those norms exclusive to Singlish), but many norms of standard English, owing to their extreme non-typicality as conditioned by Singapore’s languacultural environment, might not be adequately acquired to develop productive competence in those norms.

Interesting Quotes:

Back to Top