I have been thinking about how to link the ideas of “de-automatisation”, “foregrounding” and “metaphor” in one go. Of course, I would say some readings have an answer, e.g. Halliday’s analysis on Darwin’s Origin of Species and Newton’s Opticks (See Halliday, 1990 in Webster, 2004). But I know many of you are still stuck there: how is “foregrounding” related to “de-automatisation”?

Allow me to make a quick analogy. Language use is like a routine in life: everyone has his/her own patterns of living, as well as his/her own ways of saying things. For example, note my use of ampersands (&) in the tutorial slides. This pattern is influenced by many factors, such as my citation habits (APA conventions), the urge for space or my personal taste, etc. This is a pattern reflecting both my life routine and language habit. If you fill the calendar cells with green to show how often I use ampersands, you can automatically predict that the calendar is full of green cells.

However, there is that one day in a month I don’t use ampersands; let’s say a Saturday, and you fill that cell in red. Now, among the 31 days of October, there are 30 green cells and 1 red cell.

The red cell stands out because of the 30 green cells. We can say, the 30 green cells provide a “background” to “foreground” the red cell. We can now say, the less typical day I have on that Saturday breaks my normal pattern of using ampersands. Breaking the automatic thinking of how I perform my routine can be said as a kind of ‘de-automatisation’.

Coming back to language, de-automatisation is also about breaking the pattern of how we use language to do things. Let’s make this the case at three levels: grammar, word choice, and the overall language use. Grammatically speaking, we are accustomed to using a particular word class to construe a particular kind of meaning, e.g. verbs construe process; nouns construe things. You can see this pattern as the green cells on the calendar. However, in cases such as nominalization, the verb-process pattern breaks; verbs are nominalized to turn processes into things. You can see this “outlier” grammatical shift as the red cell. The original pattern breaks for a reason: there is a need for generalizing a set of observations of the world, especially in scientific observations, into a more stable, nominalized knowledge. This grammatical shift makes the resulting knowledge more definite and less arguable, so that we can measure it, examine it, and evaluate it. We call this shift “grammatical metaphor”, seeking a way to de-automatise the congruent, one-to-one relationship between meaning and wording. Put it simply, congruent use of grammar (verbs as processes) is the ‘background’, and the incongruent use of grammar (nominalization) is the ‘foreground’, the unexpected breaking of the pattern.

Word choice would be a simpler matter, in my opinion. In traditional notions of metaphor, the word deviates from its original meaning to refer to something else. For example, the word ‘killer’ typically refers to ‘someone who kills people’, ‘a murderer’. So the denotation (dictionary meaning) of the word is represented as the green cell. Now, if we use ‘killer’ more colloquially in conversations, we are not referring something or someone as a killer literally. Instead, it can connote a figurative meaning as ‘something/someone hard to deal with’, e.g. Eric is a killer – he’s so a*** about attendance. In this sense, the connotation of a word moves away from its original meaning, and can be seen as the red cell. Therefore, our typical habit of using ‘killer’ literally (background) breaks, turning to mean a similar concept about a thing/ person (foreground). This can be quite explicit in texts, as we are able to ‘sense’ some words don’t mean the same as we usually expect.

At a higher level, let’s say the whole text level, the functions of a text can also take “a knight’s move” for alternative purposes. For example, poetry and lyrics are the most flexible among various types of text in terms of interpretation. On one hand, a song or a poem may be a portrayal of the author’s narrative about his/her/other’s story. On the other hand, it can transcend its original purpose as a self-expressive or reflective text as the audience interpret it. Therefore, that’s why some people can seek comfort through appreciating poetry or lyrics, as they ‘free’ the meanings of the text for their own interpretations, as if it were telling their stories. To many, verbal arts are always beyond their forms (green cells, serving the function as a background) for new purposes, especially in art therapies, in that they find the definitions of the artworks for themselves, independent from their original meanings or the contexts the works are from (red, foregrounding). The way that breaks the pattern of context-text relationships can also be therefore called ‘de-automatisation’: a text is not automatically regarded as serving one particular function, but having so much potential that everyone can interpret the text on his/her own.

I attempt to summarise what I’ve said in a table. In a nutshell, what is typically expected in language use is considered as ‘backgrounding’, the expected, the unmarked choice; meanwhile, what takes its metaphorical turn to represent something else is considered as ‘foregrounding’, the unexpected, the marked choice.

	Green cell (background)	Red cell (Foreground)
Grammatical metaphor	Meaning-wording correspondence	Grammatical class shift to correspond to something else (e.g. nominalisation to represent both a thing and a process)
Conventional metaphor	Denotation (‘killer’ is a murderer, literally)	Connotation (‘killer’ is something difficult to deal with, figuratively)
Language use in general	One-to-one function or purpose of a text (a poem as a self-expressing text)	One-to-many functions of a text (a poem detours to serve therapeutic purposes, etc. to allow multiple interpretations)

To conclude, all the three cases about ‘de-automatisation’ is relevant to the notion of ‘metaphor’. The finite set of resources comprising language combine to form limitless meanings. Because these resources are limited (in terms of grammatical classes), we often find new functionalities in these resources to create new meanings.  These meanings may be random at times, but most importantly, they are finding the ‘adjacent possible’, the potentially useful function to empower us to express ourselves more economically, more vividly and more clearly.