From sense data to information

What is stored in the mind -- what it is that we 'know' -- does not, however, consist of simple 'snapshots' of environmental objects and events. Many transformations of the input data occur along the path from a stimulus object to the brain and to, ultimately, consciousness.

DEFINITION. An information-processing system is a computing system together with a program capable of executing elementary information processes on stored symbol structures. It has input-output devices that enable adequate communication with the world outside itself.

COMMENTARY. "In general, information-processing systems should contain the following components:

  1. one or more transducers which transform one for of energy into another;
  2. some elementary form of short term memory which can temporarily hold the transformed energy;
  3. a long term memory in which knowledge is built in terms of standards with which the stored inputs in the short term memory can be compared;
  4. an output system which can respond differently according to the presence or absence of a match between the received encoded information and the stored standards." (de Mey, 1982, p.24)

Figure 2 encapsulates Spradley's (1972, p.5f) account of "how sensory experience is transformed into percepts, the manner in which concepts are formed, and the way in which perceptual objects and events function as symbols."

Figure 2: Transformations of sensory information (adapted from Spradley, 1972, p.9)

Consider, by way of example, what happens when a telephone rings. It may have one or more meanings for those who hear it; but before they can act on those meanings, they must first perceive the ringing itself (object or event) -- or more exactly, perceive the pulsations of the air (channel) set in motion by the ringing. The acoustic signal that reaches the ear (sense organ) will cause a neural impulse to be sent to the brain. The resulting brain state gives rise to a percept -- a transformation of the initial acoustic event into an event in consciousness (with extraneous noise filtered out). A comparison is then made -- since percepts are unique, no two stimuli being ever totally identical in every respect -- with stored abstract images of similar events, such that the percept is recognised as an instance, or token, of a concept. It is finally in response to the concept, an abstract equivalence category[] —and not to the alarm itself—that the hearers act.

But the same object or event, in virtue of possible cultural differences between hearers, may give rise to different concepts and lead to different programs of action for different hearers. A British hearer may not, for example, recognise a ringing tone on a French or American telephone; for example, may, on dialling, believe a ringing tone to be an engaged signal. A member of a culture in which the telephone is unknown may have no corresponding concept at all. The acoustic signal may simply be heard as an anomalous sound. In brief, in learning a concept, one is learning to associate some perceptual stimulus with some cultural referent; in this case, the ringing (event) with the informational unit "someone is calling on the telephone". It is not to the sound that the hearer, in picking up the telephone, responds but to the concept:


Concepts, then, belong to the culture, not to the individual. In the first place, there is no natural (i.e., non-arbitrary) relationship between the perceptual stimulus and the concept: the same stimulus ('data') can trigger different concepts ('information') in different cultures. In the second place, for all members of the same culture a stimulus will trigger the same concept: all members 'share' the same cultural knowledge.

DEFINITION. The basic informational unit of cultural knowledge is the concept, or category. Categories simplify our world by allowing us to group objects together according to (arbitrary) salient contrastive features or attributes. The chaotic jumble of perceptual stimuli we experience day by day, minute by minute, is thus reduced to manageable terms by treating discriminably different percepts as though they were equivalent. If we could not do this, we would become hopelessly enslaved to the uniqueness of each object and event.

COMMENTARY. Most, but not all, categories of a culture have linguistic labels or names. "Language plays a vital role in what has been called the 'social construction of reality' (Berger & Luckmann, 1967). Language fixes a world that is so much more stable and coherent than what we actually see that it takes its place in our consciousness and becomes what we think we have seen" (Kress & Hodge, 1979, p.5).

The reduction of discriminably different percepts to a discrete set of cultural concepts, the elaboration and rigidification over time of conceptual 'maps' that define the cultural knowledge for some social group, and the labelling of concepts by linguistic names "not only allows us to communicate with one another but also to create conceptual worlds of the mind which are far removed from sensory experience" (Spradley, 1972a, p.8). What we 'see' becomes what our cultural knowledge has predicted we will see; those stimuli for which our cognitive maps have no corresponding concepts are either filtered out, misinterpreted, or perceived as anomalous.

Consider, by way of simple example in the context of office work (as defined by Hewitt, 1985), a standard pie graph or bar chart. The visual display is understood to represent, in highly stylised form, some collection of facts about the real world. We have learned public conventions, in reading such visual display forms, of interpreting the world in terms of n-dimensional co-ordinates where each dimension is understood to correspond paradigmatically to some scalar category (time, cost, volume, or whatever) or to some closed set of discrete entities (employees, product lines, ...). Which categories are selected determine what part of the world we are looking at and what the salient entities are. It may turn out that, to all intents and purposes, the representation supplants the world of which it is a partial image and 'becomes' that world; the content of the visual display becomes the conceptual currency of those whose business it is to generate and read such displays. Clearly, organisational work in which graphs or charts are used as an information resource—i.e., are used to mediate between members of the organisation meanings which will inform subsequent decision-making—must depend on members sharing a common ontology of objects and relations, and common conventions for the interpretation of the formal and substantial symbols that it presents.

The man in the street is a naive realist who lives in a world he can count on, a world he believes is much the same for everyone else. This view may give him a sense of security, but it must be rejected on the basis of contrary evidence. Studies in psychology have revealed that what we know of reality is constructed during the complex processes of perception and cognition. Cross-cultural studies have shown that the way people conceive of the universe differs from one society to another, and that what we take for reality is socially constructed. (Spradley, 1972a, p.8; my italics).

Figure 3: Cultural knowledge is self-reinforcing and self-confirming.

[] "When we note a common response to an array of objects which are discriminably different, we infer that the organism has formed a concept" (Spradley, 1972, p.11; cf Bruner et al, 1956, p.@)
[Back to text]