Propositional Versus Constellatory Constructs

К оглавлению
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 
119 120 121 122 123 124 125 126 127 128 129 130 131 

Propositional constructs are those, often peripheral, constructs that offer

simple descriptions of basic and, at first glance, superficial element

characteristics. A grid with the topic of ‘people’ might have ‘male – female’;

or ‘right-handed – left-handed’ as constructs. Sometimes dealing with basic

physical characteristics and easily recognisable attributes, propositional

constructs may have a very narrow and overspecific range of convenience.

One consequence of their specificity is that they offer no basis formaking inferences

on other constructs.If you know that a car is red as opposed towhite, forexample, you

can’t readily characterise it in terms of its reliability, its performance, and so on.

Constellatory constructs, on the other hand, are those which imply the position of an

element on other constructs very strongly indeed. The construct ‘family saloon ^

sports car’ is fairly constellatory for many people ^ there’s a lot wrapped up in that

distinction.Construe a car asmore of a‘family’car, and you’re prepared to construe it

on other constructs such as ‘safe ^ unsafe’,‘steady ^ exciting’,‘slow to accelerate to

60 ^ fast to accelerate to 60’, and so on; construe a carasmore ofa sports car, and the

chances are that you’d also be prepared to construe it along those dimensions (at the

other end of each!). Consequently, cars of either kind merit different marketing

strategies, and different forms of advertising based on the relevant associations to

each, as a result. In contrast, whether a car is red or white is likely to carry few such

associations. (Thisis anempirical issue, though, andmuchmarket research consists

in seeing what kinds of associations people might have for constructs which at first

glance appear to bemerely propositional.)

Also, it’s worth noting that a construct is constellatory or propositional depending on

the topic.‘Red as opposed to white’ may not be constellatory for cars, but, for some

interviewees at least, it is when it’s applied to people ^ in terms of whether they’re a

redhead ora blond(e).

Which reminds me: constellatory constructs are often characteristic of stereotyped


As with all construct characterisation, your analysis procedure is fourfold:

(1) Identification – though this isn’t necessarily clear-cut. Just keep the

propositional–constellatory distinction in mind as you look at the constructs,

thinking back to the interview process and remembering whether you agreed

to exclude some propositional constructs because you both felt they were


(2) Assess proportion. Are there a lot of constructs which are clearly

propositional or clearly constellatory?

(3) Ascribe significance in context. If you decide that there are a lot of either

propositional or constellatory constructs, it may lead you to decide that the

grid as a whole never got beyond the superficial. Perhaps there are a lot of

propositional constructs because the interviewee wasn’t encouraged to

express any of the deeper, original, or more thoughtful constructs in his or

her repertoire. On the other hand, maybe the interviewee’s thinking on this

particular topic is indeed clicheґd, and there aren’t many deeper, original or

thoughtful constructs in the repertoire, and all that’s available are

constellatory constructs.

The former is unlikely to happen after you’ve gained experience in grid

technique, and know how to relax the interviewee and encourage appropriate

rumination. The latter is a finding in itself. (And of itself. Don’t assume a

person who’s clicheґd on one topic is clicheґd on others.)

(4) Assess relationships of any propositional constructs used. It is common

to agree with the interviewee that a particular propositional construct will be

used, because it will be particularly informative when you analyse the

constructs in detail. In effect, you’re using the propositional construct as you

would a supplied construct, to check out a hypothesis you have about the

relationships within the grid.

Suppose, for example, you both decided to use the propositional construct,

‘male – female’. Looking at the grid row by row, what other constructs have

ratings which are similar to the construct ‘male – female’? This would be

useful in gender studies. Constructs being used the same way as a ‘male –

female’ construct may suggest what masculinity and femininity mean to the

interviewee: what other attributes the interviewee associates with a person’s

sex. There’s more on this in Section 6.1.2 below.

Key propositional constructs which relate to the topic of the grid, or which

encapsulate a hypothesis in a research study based on repertory grid technique,

can be very useful in analysing grid material. (You may be doing a study of the

way in which left- and right-handed people are construed, for example.)

Finally, there is one important field of application for grid techniques which

works almost exclusively with propositional constructs. Grids are used in

quality-control studies in the case of products or processes in which quality is

not readily specified in physical terms, but can be determined by the subjective

judgements of a very experienced employee. For example, it can be difficult to

determine what constitutes ‘good finish’ as distinct from ‘poor finish’ in the

case of a garment; or ‘mature taste’ as distinct from ‘immature taste’ in the case

of a cheese. One can’t tell, from these particular constructs, what the other

constructs are: in a sense, the constructs aren’t constellatory enough!

Experienced employees would be encouraged to provide as many distinct

constructs, relating to physical and measurable attributes, as they could think

of. Further analysis would show which attributes do, and which don’t, match –

with the notion of ‘good finish – poor finish’ in the case of the garment

example. Once they were explicitly identified, they could be highlighted in the

training of new quality controllers, or production employees could be trained

to recognise them during the manufacturing process as part of a zero-defects

initiative. (And, once they’re explicitly known, it may be possible to develop

physical measures for some of the attributes. For example, it turns out that

acidity is a measure of maturity in some kinds of cheese, and the acidity of

cheese can be measured more reliably in a test tube than by popping a piece

into your mouth!)