Artefacts and Family Resemblance
I develop in this paper a conception of artefacts based on L. Wittgenstein’s idea of family resemblance. My approach peruses the notion of frame, which was invented in cognitive psychology as an operationisable extension of this philosophical idea. Following the metaphor of life-cycle I show how this schematic notion of frame may be filled with the content relevant for artefacts if we consider them from the point of view of their histories. The resulting conception of artefacts provides a new insight into the current debate on artefact categorisation.
KeywordsProduction Phase Movement Mechanism Design Attribute Structural Invariant Family Resemblance
In the mid of the 1970s Nicholas Griffin made a sagacious observation on the fate of the idea of family resemblance. The gist of it has it that this idea had remained a philosophical programme without any realisation. No philosopher had bothered to employ it in order to analyse or describe any concept (Griffin 1974, p. 651). To my best knowledge, this finding has remained valid up to the present day provided that its scope is limited to philosophy. The conception of family resemblance has been quite popular, much debated, usually opposed to, but rarely, if ever, employed by philosophers. In my paper I intend to pick up this gauntlet for the (good) case of artefacts. Namely, I will build a concept of artefact in accordance with a particular interpretation of Wittgenstein’s idea and show how this concept may shed a new light on the problem of artefact categorisation.
Section 1 outlines the transition from the philosophical idea of family resemblance to its empirical validation in cognitive psychology. In particular, I sketch there the main features of L. Barsalou’s model of human concepts, which hinges upon the notion of frame. The next section employs this notion for the sake of the proper description of artefacts as objects-with-history. Section 3 shows the consequences of this description to the problem of artefact categorisation. The meta-theoretical status of my proposal is discussed in the last section.
1 Family Resemblance: from a Philosophical Idea to a Psychological Experiment
Consider for example the proceedings we call games […] if you look at them you will not see something that is common to all, but similarities, relationships, and a whole series of them at that […] look for example at board games with their multifarious relationships. Now pass to card-games; here you find many correspondences to the first group, but many common features drop out, and others appear. When we pass next to ball games, much that is common is retained, but much is lost. Are they all amusing? Compare chess with noughts and crosses. Or is there always winning and losing, or competition between players? In ball games there is winning and losing, but when a child throws his ball at the wall and catches it again, this feature has disappeared […]. (Wittgenstein 1953)
(a) The tendency to look for something in common to all the entities which we commonly subsume under a general term. We are inclined to think that there must be something in common to all games, say, and that this common property is the justification for applying the general term “game” to the various games; whereas games form a family the members of which have family likenesses. […] The idea of a general concept being a common property of its particular instances connects up with other primitive, too simple, ideas of the structure of language. It is comparable to the idea that properties are ingredients of the things which have the properties; e.g., that beauty is an ingredient of all beautiful things as alcohol is of beer and wine, and that we therefore could have pure beauty, unadulterated by anything that is beautiful. (b) There is a tendency rooted in our usual forms of expression, to think that the man who has learnt to understand a general term, say, the term “leaf”, has thereby come to possess a kind of general picture of a leaf, as opposed to pictures of particular leaves. [.…] we are inclined to think that the general idea of a leaf is something like a visual image, but one which only contains what is common to all leaves. (Galtonian composite photograph.) This again is connected with the idea that the meaning of a word is an image, or a thing correlated to the word. (Wittgenstein 1958)
object a has B, C, D, and E;
object b has A, C, D, and E;
object c has A, B, D, and E;
object d has A, B, C, and E;
object e has A, B, C, and D.
Here we can already see how natural and how proper it might be to apply the same word to a number of objects between which there is no common feature. And if we confine our attention to any arbitrarily selected four of these objects, say e d c a, then although they all happen to have B in common, it is clear that it is not in virtue of the presence of B that they are all rightly called by the same name. Even if the actual instances were indefinitely numerous, and they all happened to have one or more of the features in common, it would not be in virtue of the presence of the common feature or features that they would all be rightly called by the same name, since the name also applies to possible instances that lack the feature or features. (Bambrough 1961, p. 209–210)
Incidentally, Bambrough’s exposition allows also for an alternative interpretation of the negative part of Wittgensteinian conception: (at least) some concepts do not have the so-called characteristic intension – cf. (Ajdukiewicz 1974, p. 43–44) – i.e., if Φ is such a concept, then there exists no such set X of properties that (i) each object that falls under Φ has all properties from X and (ii) only objects that fall under Φ have all of them.
There is a tenacious tradition of thought in philosophy and psychology which assumes that items can bear a categorical relationship to each other only by means of the possession of common criteria attributes. The present study is an empirical confirmation of Wittgenstein’s (1953) argument that formal criteria are neither a logical nor psychological necessity; the categorical relationship in categories which do not appear to possess criterial attributes, such as those used in the present study, can be understood in terms of the principle of family resemblance. (Rosch and Mervis 1975, p. 603)
The hypothesis tested in (Rosch and Mervis 1975) can be summarised as follows. The subjects of their experiments were confronted with six classes of concepts with respect to which they were asked to make prototypicality judgments and/or describe these concepts by means of lists of attributes. The classes in question contain both “natural” categories, which correspond to the semantics of English words and phrases, and artificial categories, which were meaningless strings of letters and digits. Suppose that we define the similarity relation between such concepts in terms of overlap between those attributes: one concept is similar to another to the degree of the overlap between the attributes they have. The first part of the hypothesis in question posits that one concept is prototypical of another super-ordinate concept in proportion to the extent the former bears such similarity to other members of the latter. For instance, “chair” is a prototypical member of “furniture” to the extent the attributes of “chair” overlap with the attributes of other members of “furniture” such as “sofa” or “table”. The second part of the hypothesis has it that the most prototypical members have least similarity to concepts in other super-ordinate concept. Thus, the the attribute of “chair” should share less attributes with, say, “weapon” then, for example, with “sofa”. The six experiments described in (Rosch and Mervis 1975) proved these claims to be valid. As the reader may easily verify (Rosch and Mervis 1975) implicitly presuppose Wittgenstein’s notion of family resemblance along the lines established by Bambrough’s interpretation.
The results obtained by E. Rosch and her associates invigorated the research on the structure and dynamics of our conceptual life. Besides a number of specific experimental results three main types of theoretical models of concepts emerged: (i) prototype view, (ii) exemplar view, and (iii) theory view (Murphy 2002, ch. 3). These three models are theories with different predictive and explanatory capabilities, still they share some salient characteristics. Barsalou (1991) argues that there exists a distinct conceptual framework that unifies the main theories of concepts and whose capabilities are commensurable to the complexity of human knowledge. The main component of his proposal is the notion of frame, which is claimed to constitute a reasonable generalisation of the notion of attribute list.
Another specific feature of frames is their extendibility. Since (i) frames represent concepts and (ii) both attributes and their values are concepts, therefore there may exist frames that represent other frames’ attributes and values. Furthermore, there are frames for structural invariants and value constraints.5 Metaphorically speaking, there might be frames within frames.
Consider the word “newspaper” and note which of its features come to mind. Now consider the word “newspaper” in the context of building a fire. Whereas feature flammable probably didn’t come to mind when you consider “newspaper” in isolation, it probably did when you considered it in the context of building a fire. Many researchers have implemented such manipulations in experiments and observed large effects on verification time. (Barsalou 1993, p. 31)
Formally, the main frame for a concept can be represented as a directed connected labelled graph, with labels both on vertices and edges, and its variants as its connected subgraphs – see (Petersen 2007).
Barsalou (1991) shows how this conception of frames may be employed as a common conceptual framework for the classical view of concepts and its main non-classical competitors. Obviously, each of these views will use frames in a different way to express the internal structure of concepts, their role in categorisation, etc. In particular, the prototypicality predictions will differ among those views even if they are expressed by means of frames.
It is the idea of Barsalou frames that has shaped my proposal of a family resemblance notion of artefact.
2 A Family Resemblance Notion of Artefact
Concept – when the requirements for the artefact to be designed are defined and analysed;
Development – when the artefact is designed;
Production – when the artefact comes into existence;
Utilisation – when the artefact is used;
Support – when the artefact is serviced;
Disposal – when the artefact is retired, archived or disposed.
Technology is not typology. It takes into account the entire lithic material without preferentially isolating what we choose arbitrarily to call “tools”. It places each item in the sequence of technical actions beginning (after its conception and prior contemplation) with the raw material and ending with the abandonment, the “death” of the tool assemblage. Even when fragmented into thousands of microliths and “debris”, a lithic assemblage always forms a coherent whole bound together by a methodical scheme. (This is an English translation from (Bar-Yosef and Peer 2009) of one of the opening passages in (Tixier et al. 1980))
These five processes will be interpreted as attributes of frame representing artefacts. I will develop this idea in two stages: first introducing basic and then extended frames.
2.1 Basic Frames
Strictly speaking, basic (and also extended) frames directly represent the histories of artefacts and artefacts are represented indirectly as object participating in those histories. Thus, strictly speaking, the five attributes of design, production, use, service, and disposal are attributes of the histories of artefacts and not of the artefacts themselves. The latter are characterised by the attribute of being designed, being produced, etc. Consequently, all my claims regarding frames representing artefacts are to be interpreted in this manner.
In principle any connected subgraph of the main basic frame may count as a variant of this frame, however, there are certain philosophical limitations on this combinatorial variety, i.e., not every subgraph of the main basic frame represents an artefact. However brushing aside, for a moment, these “excluded subgraphs”, one can say that an artefact is an entity that was either designed or produced or used or serviced (maintained) or disposed.
Let me now introduce a bit of notation to speak about main basic frames and their subgraphs. A frame subgraph will be represented as a sequence of acronyms of its attributes. The order of this sequence follows the order of attributes in the enumeration above because the basic frames do not carry any information about the temporal order of their attributes. This implies, among other things, than no sequence contains more than one occurrence of the same attribute abbreviation. Thus, for instance, DePrUse will represent a variant of the basic main frame, which represents those artefacts that are/were designed, produced, and used. Each such sequence is meant to provide a complete characteristics of a given artefact type, so DePrUse represents those artefacts that are/were not serviced or disposed. This implies that the categories I define in terms of frames are pairwise disjoint, e.g., there is no artefact that both DePrUse and DeUseSerDi represent.
Now consider the class of subgraphs of the main basic frame that contain the design attribute, but do not contain the production attribute. Consider for example the “would-be” frame DeUse. It is to represent those objects that were designed and used, but were not produced, i.e., which did not make it to existence. The problem with them is obvious: you cannot use an object unless it exists and the frame DeUse pretends to represent such objects. Consequently, the following subgraphs of the main basic frames are not considered here as its variants: DeUseSerDi, DeUseSer, DeUseDi, DeUse, DeSerDi, DeSer, and DeDi.
Among the remaining subgraphs there are three other disputable cases: SerDi, Ser, and Di. Consider the first case. It represents those objects that are somehow being serviced, i.e., maintained in a working condition, and disposed without being designed, produced or used. Since they were not produced, only the so-called natural objects like (some) stones and (some) trunks may fall under SerDi. Suppose that a certain stone is of this kind. If this stone were an artefact of the type UseSerDi, we could make up a story that someone found it, used it as an altar for several years, during these years he removed moss and lichen, and finally disposed it during a crisis of faith. But what kind of story can be told when we want to describe an object that was only serviced and then disposed? The main problem here is how to make sense of the operation of maintenance. If the object in question is neither designed nor produced nor used, how can we meaningfully claim that we maintain it? If it were designed and produced, we could claim that maintenance means restoring some of its designed and/or produced features. If it were only used, we could claim maintenance means restoring some of its properties that are important for the way we use it. Without either design or production or use one cannot explain and justify the claim that a certain process is a maintenance operation with respect to an instance of SerDi. A similar argument disqualifies Ser as a variant of the main basic frame.
The case of Di artefacts is different. Now the problem is that if Di were a frame for artefacts, then any object we intentionally destroy would become an artefact just because it was destroyed thereby. One could even argue that it becomes an artefact at the end of its disposal, i.e., it becomes an artefact when it ceases to exist. These consequences seem to overstrain our common-sense intuitions concerning artefacts or even objects in general.
Primary artefacts are ones used directly in production, such as axes, needles, bowls, etc. Secondary artefacts are internal and external representations of primary artefacts, and they are created and used “in the preservation and transmission of the acquired skills or modes of action or praxis by which this production is carried out” (p. 202). As such, secondary artefacts are representations of these modes of action. Representing a mode of action also means that these artefacts are related to conventions, as in rules and norms. Tertiary artefacts are imaginary artefacts such as art or free play or game activity. Such artefacts have lost their original role of representation, since they have become “abstracted from their use in productive praxis” and from their “direct representational function”. (Susi 2006, p. 2211)
If we agree that Wartofsky’s “tertiary artefacts” are bona fide artefacts, then among all frames only the De may extend our notion of artefacts to include them. After all, most of the tertiary artefacts are designed in such a way not to make it to the existence. Similarly, among the “secondary artefacts” there are those that can be represented only by this type of frames. For instance, the category of such entities as plans (and also strategies, methods and the like) include both plans that were implemented and those that were never realised. It seems that the former do not differ, qua plans, from the latter. And again only De frame may represent the latter. Needless to say, artefacts represented by De are borderline cases of artefactuality, but the very idea of family resemblance is to include also such atypical instances.
As a result, we are left with 21 variants of the main basic frame. This frame and its variants constitute the “abridged version” of my family resemblance notion of artefact. Although this notion is extensionally capacious, including as many borderline cases of artefacts as possible, it does not include all of them. If you use a tiled stove to heat your apartment, then the ash and the soot it produces are not artefacts in the sense of this conception. On the other hand, when using this stove you also intend to obtain a large quantity of ash, then the ash produced thereby becomes an artefact. Generally speaking, the view elaborated in this paper implies that no object that is an unintended result (of a process of some kind) is an artefact.
2.2 Extended Frames
Let me now expand the notion of basic frames by elaborating the specific structure and content of each attribute: De, Pr, Use, Ser, and Di. Since a proper characterisation of each of them would, in fact, require, a separate paper, my characterisation will be rather cursory.
Each attribute is construed here as (a representation of) a process (or a perdurant, if you like) (i) in which some agent or agents actively participate(s) as its actor(s) and (ii) whose beginning and end are characterised in terms of certain objects (i.e., endurants) and properties.8 The latter will be referred to as “input” and “output” objects and properties to a process in question.9
Besides the “usual” ontological difference between objects and their properties, I will employ an additional distinction. Namely, an object that is an input to a process is assumed to exist (at least) at the moment at which the process starts. Similarly, for output objects: an object that is an output to a process must exist (at least) at the moment at which the process ends. On the other hand, both input and output properties (to a process) carry no existential commitment. More precisely speaking, a property at an input (resp. output) to a process either exists at the beginning (end) of the process, i.e., is exemplified at that time by some object, or is an object of someone’s intention (at that time), i.e., is being desired or intended (by someone) to exist. For instance, when a customer specifies his or her requirements, I will represent these requirements as a set of intended, but unexemplified properties.10
Any production process culminates in the creation of a new object. I assume that this event goes beyond a simple modification of existing objects and genuinely increases their number. The “input” properties to this process are those that the producer of the artefact, i.e., this agent who actively participates in production, intends to exemplify in the artefact. The output properties correspond to those “input” properties that are successfully exemplified at the end of this stage. Thus, for each property from the output set, which is actually exemplified by the artefact, there exists an “intended”, unexemplified property in the input set. Since I want to keep the ontological commitments of my conception minimal, I do not claim that the output properties are those “input” properties that are successfully exemplified. I do claim that there exists an injective mapping from the output set to the input set: for each “exemplified” property from the former set there exists exactly one property from the latter set that was intended to be exemplified. If you believe that there are unexemplified properties, which simply get exemplified at some stage in their existence, you could say that this mapping is the relation of identity. In any case it follows then a process is an act of production only if the producer managed to exemplify at its end at least property he or she intended to at its beginning.11
Finally, let me emphasise that a process of production need not be preceded by a design phase. When a production agent starts his or her activity without considering any goals or rationale for the artefact to be produced, but he or she simply intends that a certain set of properties should be exemplified and this intention is not a result of an earlier process of deliberation, then this situation is a case of a life-cycle that starts at the production phase. Obviously, it is possible that this production phase is the only phase in the life-cycle – this happens when the produced artefact is “abandoned”, i.e., when nobody will use, maintain, and dispose it.
An act of use is characterised (i) by means of an object being selected as a tool or instrument and (ii) by means of its properties because of which it is selected as such. Obviously, the object is the artefact itself, which exemplifies the selected properties. I assume that the selection of the object and its properties happen at the beginning of each use phase.
There is a clear difference between design and production, on the one hand, and use, on the other. The difference at stake amounts to whether a phase may occur more than once in a life-time of one artefact. If you assume the principle to the effect that no entity can have two beginnings in existence (see the famous quote in (Locke 1689, II.27.1)), then this principle guarantees the uniqueness of the production attribute within a single frame for one artefact. The case of the design attribute is slightly different. In principle it may happen that this phase is composed of a number of separate activities that do not overlap either in space or time. However, since my artefact frames are not complex enough to account for such differences, e.g., since they do not represent time, I will coalesce all such separate events into one design process.
Similar remarks can be made about the two subsequent attributes: Ser and Di. The former represents all cases of service activity that the artefact undergoes during its lifetime. The latter represents an event of its destruction, which is unique once you accept the counterpart of Locke’s principle: no entity can have two ends in existence. Incidentally, let me note that I do not exclude the possibility that the object has been destroyed during its use.
It is quite straightforward that use events are different from production phases in that when a use event for an artefact x commences, x exists, while when a production phase (for x) is being initiated, x does not exist yet. Thus, the difference between, say, an artefact represented by the frame Use and an artefact of the frame PrUse is that the former (e.g., a stone) comes to existence without human intervention, but happens to be used, and the latter (e.g., a sculpture that is created “on the fly”) is a result of some intentional production process, which result at some later stage of its life time is also used. So, when you wrap a piece of carp fish in a banana leaf, this leaf is an instance of Use, but when you build a food container out of banana leaves and then use it as a kitchen utensil, then this container is not an instance of Use but of PrUse.
a subset of the De output set of properties;
a pseudo-subset of the Pr output set of properties;
a pseudo-subset of the set of properties that were selected by a user of this artefact at the beginning of one of its earlier events of use.
that the artefact ceases to exist and
that some other objects in its environment exemplify certain properties.
The first type of disposal attribute (Fig. 10a) covers those situations in which the disposal agent attempts to create a new environment after the artefact ceases to exists. Obviously, the output set of properties (of this attribute) would then be a pseudo-subset of the set of input properties. The second type of disposal attribute (Fig. 10b) covers those situations in which the disposal agent attempts to retain some aspects in the artefact’s environment. Now, the output set of properties would be a subset of the set of input properties. The third type, probably the most common, combines the previous cases.
for each frame that contains the Di attribute, its three variants that correspond to the three cases of the disposal activity depicted in Fig. 10;
- 2.for each frame that contains both De and Pr, its two variants that correspond to two possibilities:
the input set of properties from Pr is a subset of the output set of properties from De;
the input set of properties from Pr is not a subset of the output set of properties from De;
- 3.for each frame that contains both De and Di, its two variants that correspond to two possibilities:
the input set of properties from Di is neither a subset nor a pseudo-subset of the input set of properties from De.
The relations of being a subset and of being a pseudo-subset are considered here as the structural invariants of the extended frames. Moreover, those that are mentioned in the above enumeration are considered facultative, i.e., they may not occur in some extended frames.
3 Multi-faceted Categorisation of Artefacts
Does the above conception shed any new light on the problem of artefact categorisation?
Family resemblance across concepts of artefacts
\(\subseteq \) – when the extension of a basic frame is a subset of the extension of a respective concept;
\(\cap \) – when the extension of a basic frame properly overlaps the extension of a respective concept;
\(\emptyset \) – when the extension of a basic frame does not overlap the extension of a respective concept.
the conditions listed in (Baker 2007, p. 52–53) imply that the history of an artefact always involves an event of creation of a new object, i.e., the artefact itself. This event may be preceded by an earlier design activity and may be followed by subsequent events of use, service, and disposal;
that each event that is represented by De, Pr, or Use is an event of intentional selection (IntentionalSel) in the sense of (Borgo and Vieu 2009);
that the notion of modification used in (Dipert 1993) subsumes my Pr attribute and that not every artefact that falls under one of my basic frames exhibits “modified properties [that] were intended by the agent to be recognized by an agent at a later time as having been intentionally altered for that, or some other, use” (Dipert 1993, p. 29–30);
that (Hilpinen 1993) makes a distinction between an event of selection of desirable properties (my De attribute) and an event of implementation of those properties (my Pr attribute) and that the history of an artefact requires both;
that the notion of design employed in (Borgo 2011, p. 6–8) by R. Mizoguchi and Y. Kitamura corresponds to my De frame and that their notion of production corresponds to my Pr frame;
that the so-called “dual definition” from (Houkes and Vermaas 2010, p. 158–160) implies that a fully-fledged artefact both undergoes a proper production process and is involved in one or more use events and that there are such use events that do not fall under their notion of use plan.
Baker (2007) selects the production phase as the essential aspect of artefacts;
Borgo and Vieu (2009) select any phase without differentiating between them – with one exception: De frame does not represent artefacts in their sense
Dipert (1993) selects use as the essential aspect;
Hilpinen (1993) selects the phases of design and production;
R. Mizoguchi and his associates also select design and production;
Houkes and Vermaas (2010) select production and use.
Secondly, the family resemblance notion provides a new set of terms for the debate about artefact categorisation. The current state of this debate is dominated by the “categorisation-by-function” camp and its opponents. According to the former camp whether a certain artefact belongs to a given category of artefacts depends solely on its functions (e.g., (Baker 2007)). The dissenters to this view, while acknowledging a role of functions in artefact categorisation, point out to other aspects, like the intentions of the artefact’s designer (Vaesen and van Amerongen 2008) or a use plan for this artefact that was envisaged during its design – see (Houkes and Vermaas 2010).
- 1.its properties at the design phase:
those that come out of this phase as its intended properties;
- 2.its properties at the production phase:
those that are planned to be exemplified
those that are actually exemplified;
- 3.its properties at one of its use phases:
those that a user of the artefact considers at this phase while he or she selects it as a tool;
- 4.its properties at one of its service phases:
those that are planned to be restored (re-exemplified) at this phase
those that are actually restored after this phase is finished.
those properties that constitute the rationale for the creation of the artefact at the beginning of the design phase (the “input” properties of the design attribute);
those properties that the disposal agent intends either to establish or to maintain in the artefact’s environment at the beginning of the disposal phase;
those properties that are actually established or maintained in the artefact’s environment at its disposal.
2a being reduced to 1a
4a being reduced to 2a or to 3a
6 being reduced to 5
- 1.intended properties
“output” properties of De
“input” properties of De
- 2.exemplified properties
“output” properties if Pr
properties considered by a user (of an artefact) at a use phase
properties actually restored after a service phase is finished
“output” properties of Di.
The above conception introduces a new perspective on the problem of artefact categorisation. Instead of analysing the ontological status of artefacts as if they did not evolve in time or instead of focusing at one, allegedly significant, phase in their history, I recommend tracing the changes of this status throughout their life-cycles. This perspective brings forward two groups of properties that were neglected in the debate on artefact categorisation. The first group involves those properties that are either intended to be restored or are actually restored in an event of artefact maintenance. The other type concerns those properties that are either restored or established during the disposal of the artefact. So instead of classifying artefacts with respect to their designed properties one can classify them with respect to those properties that are important from the point of view of maintenance or disposal. At the same time the perspective undermines the significance of the functions vs non-functions distinction in this debate. Although the most appropriate “place” for artefactual functions appears to be at the output of the design phase, one can find examples of artefacts whose functions come into existence at the input of the production phase or at the input of a use or service event. The former cases concern those situations in which the production specification diverges from the design specification, e.g., when a manufacturing agent decides that it is not feasible to implement one of the designed properties or that it is feasible and desirable to implement a property that was not envisaged in the design. On the other hand, when you use a certain artefact to achieve a goal that is not among its goals defined in the design or production phase, you are involved in a situation of the latter kind. So when you use a hammer as a paperweight, a new function of this hammer comes into existence.
My account of artefacts represents them by means of a family resemblance concept. Even if some details do not correspond to how we actually conceptualise and categorise them, the psychological research on concepts and categorisation makes it improbable that the “real” concept of artefacts may be captured by means of a unique set of necessary and sufficient conditions. The classical approach to concepts seems to be closed for those who intend to engage rather in descriptive than in prescriptive philosophy. The device of family resemblance interpreted as a frame in the sense of L. Barsalou brings my proposal closer to the actual activity of artefact categorisation.
[…] while modifying the original vague meaning of a word, sharply delineates its extension, but takes into consideration the original vague boundary of its extension, as suggested by the original meaning. (Ajdukiewicz 1974, p. 75)
My aim was thus not to represent our concept of artefacts in all detail. First, the linguistic variety of the term “artefact” and its derivatives13 and the differences between philosophical theories of artefacts cast a doubt whether there is a single concept that covers the whole domain of artefacts. Even if it turns out that there is, then it seems that the concept in question, as opposed to the so-called basic categories, is too abstract and too vague to be stable both across different subjects who use it and different times when it is used.
Being located between the two extremes of the descriptive and the prescriptive analysis my conception is to provide a new methodological perspective on the domain of artefacts with respect to the problem of adequate characterisation of its members. The novelty in question concerns the fact that instead of propounding a single, crisp notion of artefact, I provide a family of such notions, which can be applicable in various conceptual circumstances. In particular my conception is capable of accommodating different epistemological interests of those who aim to categorise artefacts. As the research surveyed in (Malt and Sloman 2007) attests, an event of artefact categorisation usually depends on the cognitive task to which it is assigned. This dependence involves, among other things, an appropriate selection of those properties that are relevant for this task and which are thus the basis for this categorisation event. My conception attempts to delineate the range of possible selections and to classify them. Moreover, it propounds two “new” types of properties suitable for this purpose. In particular, the “disposal” properties seem to be a realistic alternative to the traditional bases of artefact categorisation.
Nevertheless, experimental validation of the above model would be both a stimulating and instructive challenge if it did not exceed the research competencies of the author of this epistemic artefact.
M. Pelczar labels in (Pelczar 2000) this interpretation as standard.
The idea of family resemblance was also tested for applicability in Artificial Intelligence. (Carrara et al. 2011) attempt to show why the notion of engineering function may be represented as a family resemblance notion and what consequences this claim may bring for the formalisation strategies in applied ontology.
Barsalou represents an attribute of a concept by means of a single line that is labelled “aspect”.
As a matter of fact, Barsalou speaks about constraints in general and draws two orthogonal partitions thereof: (i) attribute constraints vs value constraints (ii) contextual constraints vs optimisations. Now, from the extensional point of view the notion of attribute constraint is subsumed by the notion of structural invariant since both the latter and the former express certain relationships between frames’ attributes even though their extra-logical properties are different, e.g.: “rather than being normative [as structural invariants are], constraints produce systematic variability in attribute values.” (Barsalou 1992, p. 37). Therefore, I mention only value constraints as a frame component.
Barsalou provides us with the example of the parthood relation, which is a structural invariant in great deal of frames. Following (Winston et al. 1987) he claims that parthood can be represented by a frame with four attributes: functionality, separability, homeomeronymy, and spatio-temporal extent.
The term “the life (of an object)” is to denote the mereological sum of all processes and events in which this object participates, i.e., its history.
Since each attribute has “the same” set of values, I do not include them in this figure for the sake of readability. For the same reason I drop the “aspect” label I used before.
Although I acknowledge the possibility that numerous agents may participate in such process (either on a cooperative or competitive basis), for the sake of simplicity I will use the singular form to refer to them and their activities.
I use the verb “exemplify” to denote the relation of exemplification as defined in (Lowe 2006, ch. 2).
If one property is intended to be exemplified and the other is “the same” property but exemplified, I will say that they are equivalent. If two sets X and Y of properties are such that for each property in X, there exists an equivalent property in Y, I will say that X is a pseudo-subset of the Y.
For example, the Directive 2000/53/EC of the European Parliament and of the Council of 18 September 2000 on end-of life vehicles specifies the requirements on the sites for storage and treatment of end-of-life vehicles and the sequence of treatment operations. These requirements are to guarantee that such end-of-life vehicle fluids as fuel, motor oil, gearbox oil, transmission oil, hydraulic oil, cooling liquids, antifreeze, brake fluids, battery acids, air-conditioning system fluids are separated, collected, and secured in impermeable containers so that the level of contamination of the soil, water, and atmosphere in the vicinity of these sites is kept at a minimum level. In this case the environment for a car to be scrapped is constituted by the storage and treatment site and the vicinity of this site.
The reader may find ample textual evidence thereof in (Parker 2010).
This notion of modification is sufficiently broad to include also “negative modifications”, i.e., those cases when some properties of a given object were deliberately “left alone”.
Although some of these conditions depend on the others, I list all of them to show the reader the theoretical alternatives which Hilpinen’s conception makes room for.
- Ajdukiewicz, K. 1974. Pragmatic Logic. D. Reidel Publishing Company.Google Scholar
- Bambrough, R. 1961. Universals and family resemblance. Proceedings of the Aristotelian Society 61: 207–222.Google Scholar
- Barsalou, L.W. 1991. Components of conceptual representations: From feature lists to complex frames. In Categories and concepts: Theoretical views and inductive data analysis, eds. I. van Mechelen, J. Hampton, R. Michalski, P. Theuns, 97–144. San Diego: Academic Press.Google Scholar
- Barsalou, L.W. 1992. Frames, concepts, and conceptual fields. In Frames, fields, and contrasts, eds. A. Lehrer, E. F. Kittay, 21–74. Lawrence Erlbaum Associates.Google Scholar
- Barsalou, L.W. 1993. Flexibility, structure, and linguistic vagary in concepts: Manifestations of a compositional system of perceptual symbols. In Theories of memory, eds. A. Collins, S. Gathercole, M. Conway, 29–101. London: Lawrence Erlbaum Associates.Google Scholar
- Barsalou, L.W., W. Yeh, B.J. Luka, K.L. Olseth, K.S. Mix, and L.-L. Wu . 1993. Concepts and meaning. In Chicago linguistics society 29: Papers from the parasession on conceptual representations, eds. K. Beals, G. Cooke, D. Kathman, K. McCullough, S. Kita, D. Testen, 23–61. Chicago: Chicago Linguistics Society.Google Scholar
- Borgo, S., and L. Vieu. 2009. Artefacts in formal ontology. In Handbook of the philosophy of science, ed. A. Meijers, 273–307. Elsevier.Google Scholar
- Borgo, S., M. Franssen, P. Garbacz, Y. Kitamura, R. Mizoguchi, and P.E. Vermaas. 2011. Technical Artifact: An integrated perspective. In Formal ontologies meet industry, eds. P. Vermaas, V. Dignum, 3–15. IOS Press.Google Scholar
- Carrara, M., and P. Garbacz, P.E Vermaas. 2011. If engineering function is a family resemblance concept: Assessing three formalization strategies. Applied Ontology 6: 141–163.Google Scholar
- Dipert, R. 1993. Artifacts, art works and agency. Philadelphia: Temple University Press.Google Scholar
- Garbacz, P. 2009. What is an Artefact Design? Techné 13(2): 137–149.Google Scholar
- Goodman, N. 1955. Fact, fiction, and forecast. Cambridge: Harvard University Press.Google Scholar
- Griffin, N. 1974. Wittgenstein, universals and family resemblances. Canadian Journal of Philosophy 3(4): 635–651.Google Scholar
- Hilpinen, R. 1993. Authors and artifacts. Proceedings of the Aristotelian Society 93: 155–178.Google Scholar
- Houkes, W., and P.E. Vermaas. 2010. Technical functions: On the use and design of artefacts. Springer.Google Scholar
- Locke, J. 1689. An essay concerning human understanding. Thomas Bassett.Google Scholar
- Lowe, E.J. 2006. The four-category ontology. A methaphysical foundation for natural science. Oxford: Clarendon Press.Google Scholar
- Malt, B.C., and S.A. Sloman. 2007. Artifact categorization: The good, the bad, and the ugly. In Creations of the mind. Theories of artifacts and their representation, eds. E. Margolis, S. Laurence, 85–123. Oxford University Press.Google Scholar
- Murphy, G.L. 2002. The big book of concepts. Cambridge: MIT Press.Google Scholar
- Parker, P.M. ed. 2010. Artefact: Webster’s timeline history, 1148–2007. ICON Group International, Inc.Google Scholar
- Petersen, W. 2007. Formal representation of frames. The Baltic International Yearbook of Cognition, Logic and Communication 2: 151–170.Google Scholar
- Susi, T. 2006. Tools and artefacts - knowing ‘where-from’ affects their present use. In 28th annual conference of the cognitive science society in cooperation with the 5th international conference of the cognitive science society, vol 4, 2210–2215. Cognitive Science Society.Google Scholar
- Tixier, J., M.-L. Inizan, and H. Roch. 1980. Prehistoire de la Pierre Taillee. Terminologie et technologie, vol 1. CREP, Antibes.Google Scholar
- Wartofsky, M.W. 1979. Models: Representation and the scientific understanding. Dordrecht: D. Reidel Publishing Company.Google Scholar
- Wittgenstein, L. 1953. Philosophical investigations. New York: Macmillian.Google Scholar
- Wittgenstein, L. 1958. The blue and brown books. Basil Blackwell.Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.