Epistemological Significance of Models in Science
3
It has to be emphasized, however, that in order to know which of the traits
should be neglected and which retained, we inevitably require a complete
knowledge of all the features of the model. This requirement, of course,
can only be satisfied b\' artificial objects; of these we have to regard rather
arbitrarily the one as a svstem and the other as a model. Therefore, whenever
an exact comparison between models and naturally occurring originals
seems possible, this is truh an "artefact of perception" [12]. (3) The third
function of the model, its subjectivization, expresses the fact that models
cannot exist in their own right but rather that they are designed and applied
in order to satisf\' certain needs which arise only in relation to certain
subjects.
Models in Epistemology
In ralsmg the question of the position and function of models in
epistemology, we are confronted with the modern theory of cognition [7].
This theory states that science consists of two parts, linked by a third or
intermediate one. The first part, concerned with the immediate observation
of facts, represents the base and may be described in appropriate terms Lo
(language of observations) or 1.0 ' (language of measures); the second part
is the apex as represented by LT (language of theories). Between the two
parts we ma\ insert the intermediate one L c , consisting of rules of correspondence between base and apex. Thus it seems to be beyond any doubt
that models are indeed tonls for effecting such correspondence. In particular,
we may identify the object (or system) "conceived" with Lo, together
with the more or less "concretized" object (system) as introduced by
~frLLER [9]. On the basis of the general role of a model, we may now endeavor to search for an approach to LT , i.e. the "abstract" system. Since the
two (or tbree) levels of language do not express a unidirectional principle
of construction, e.g. from base to apex, but rather are more or less interdependent, we can assume that the model may be applied both to obtain
knowledge (by ascending from Lo to L T ) and to verify it (by descending
from LT to Lo). Moreover, the three main functions of a model, as compared
with the respective original, viz. mapping, reducing and subjectivization,
appear to be equalh valid in both directions.
By inference from their intermediate position (pertaining to the level
of correspondence), it becomes obvious that models represent only the
sphere of signs-provided we accept the division of "reality" into (a) a
sphere of objects, (b) a sphere of signs and (c) a semantic sphere in which
the signs attain a certain meaning [10].
Therefore, models need to be distinguished in terms of possible signs
and hence at least two t\pes may be delineated [4]: (a) iconic models,
preserving some perceptual elements, and (b) symbolic models, in which
1 *
3
It has to be emphasized, however, that in order to know which of the traits
should be neglected and which retained, we inevitably require a complete
knowledge of all the features of the model. This requirement, of course,
can only be satisfied b\' artificial objects; of these we have to regard rather
arbitrarily the one as a svstem and the other as a model. Therefore, whenever
an exact comparison between models and naturally occurring originals
seems possible, this is truh an "artefact of perception" [12]. (3) The third
function of the model, its subjectivization, expresses the fact that models
cannot exist in their own right but rather that they are designed and applied
in order to satisf\' certain needs which arise only in relation to certain
subjects.
Models in Epistemology
In ralsmg the question of the position and function of models in
epistemology, we are confronted with the modern theory of cognition [7].
This theory states that science consists of two parts, linked by a third or
intermediate one. The first part, concerned with the immediate observation
of facts, represents the base and may be described in appropriate terms Lo
(language of observations) or 1.0 ' (language of measures); the second part
is the apex as represented by LT (language of theories). Between the two
parts we ma\ insert the intermediate one L c , consisting of rules of correspondence between base and apex. Thus it seems to be beyond any doubt
that models are indeed tonls for effecting such correspondence. In particular,
we may identify the object (or system) "conceived" with Lo, together
with the more or less "concretized" object (system) as introduced by
~frLLER [9]. On the basis of the general role of a model, we may now endeavor to search for an approach to LT , i.e. the "abstract" system. Since the
two (or tbree) levels of language do not express a unidirectional principle
of construction, e.g. from base to apex, but rather are more or less interdependent, we can assume that the model may be applied both to obtain
knowledge (by ascending from Lo to L T ) and to verify it (by descending
from LT to Lo). Moreover, the three main functions of a model, as compared
with the respective original, viz. mapping, reducing and subjectivization,
appear to be equalh valid in both directions.
By inference from their intermediate position (pertaining to the level
of correspondence), it becomes obvious that models represent only the
sphere of signs-provided we accept the division of "reality" into (a) a
sphere of objects, (b) a sphere of signs and (c) a semantic sphere in which
the signs attain a certain meaning [10].
Therefore, models need to be distinguished in terms of possible signs
and hence at least two t\pes may be delineated [4]: (a) iconic models,
preserving some perceptual elements, and (b) symbolic models, in which
1 *