realized. Malfunctioning artifacts remain elusive from an ontological point of view, not least owing to the complexity of both notions of function and artifact. This paper aims to take the initial steps towards a unifying account of malfunctioning artifacts that centers around the notion of realizable entity. We begin by examining some major distinctions between theories of artifactual functions (e.g., intention-based and realizable-centered). Based on this examination, we propose to provide an ontological account of the malfunctioning of artifacts by drawing on a “realizable-centered approach to artifacts” which is recently proposed in formal ontology. The pivotal idea of this account is that an artifact is malfunctioning when some relevant “intentional realizable entity” thereof is not ∗Corresponding author.
Malfunctioning is pervasive. It figures both in the natural world (e.g., the hyperfunctioning of the heart) and in the artifactual world (e.g., the meltdown of a nuclear reactor). It is nonetheless notoriously dificult to analyze the notion of malfunctioning from an ontological viewpoint. This is not least because the term “malfunctioning” has been very diferently used in diferent domains and, more fundamentally, because there is no broad agreement about what functions are to be ontologically analyzed.
This paper aims to sketch a general theoretical framework for understanding the malfunctioning of
artifacts that is founded on the ontology of realizable entities. A realizable entity is a property that
can be realized in associated processes of a specific correlated type in which the bearer participates
— a conceptual view of realizable entities (e.g., [
In particular, we will consider the malfunctioning of artifacts by leveraging a “realizable-centered
approach to artifacts” [
CEUR Workshop
ISSN1613-0073 of artifacts in general, ranging widely from technical artifacts (e.g., screwdrivers) to artworks (e.g., ready-mades) and spiritual artifacts (e.g., amulets). It can be therefore expected to serve as a point of reference from which to compare existing diferent accounts of the malfunctioning of artifacts that are based on diferent conceptions of functions.
The paper is organized as follows. Section 2 spells out some important distinctions between theories of artifactual functions (e.g., intention-based and realizable-centered) and introduces some basic notions (e.g., intrinsic and extrinsic dispositions) used in a realizable-centered approach to artifacts. Section 3 proposes an ontological analysis of the malfunctioning of artifacts by deploying the existing realizablecentered analysis of artifacts. Section 4 briefly discusses related work. Section 5 concludes the paper. Note that we will write terms for types and particulars in italics and bold, respectively (e.g., Screwdriver and pebble1).
Section 2.1 scrutinizes three prominent distinctions between theories of artifactual functions with a focus on malfunctioning artifacts: particular-oriented and type-oriented (Section 2.1.1), intentionbased and realizable-centered (Section 2.1.2) and disposition-grounded and disposition-free (Section 2.1.3). This will specify the scope of our investigation in this paper and motivate the application of a realizable-centered approach to artifacts (Section 2.2) to understanding the malfunctioning of artifacts.
Talk of artifactual functions is frequently coupled with talk of domain-specific artifact types, e.g., the
type Screwdriver in the common statement: “This screwdriver is malfunctioning, as it is incapable of
turning screws”. For example, Jespersen & Carrara [
In contrast with such “type-oriented” theory of artifactual functions, for instance, Baker [14, p. 85] provides a “particular-oriented” view of the malfunctioning of artifacts as follows: (M) x is a malfunction of an artefact a if and only if:
(a) x is a failure to perform the intended function of a, where the intended function of a is such that it is physically possible to be performed, and
(b) x occurs when a competent operator tries to use a to perform its intended function under conditions for which a was designed.
Note that Baker uses italicized letters as individual variables, whereas we use letters in boldface as individual constants. According to Baker’s formulation (M), it is a particular artifact (“a”) that is said to malfunction or not.
In this paper we will adopt a particular-oriented theory of artifactual functions. For one thing, domain-specific artifact kinds may be certainly crucial to account for function ascriptions, i.e., roughly, for sentences or propositions that describe the entities and phenomena involving functions. However, the type-oriented theory of artifactual functions may not serve to theoretically underwrite a general ontology of functions and (mal)functioning as clearly as the particular-oriented theory can, because dealing with domain-specific artifact kinds may need careful ontological consideration.
To illuminate this point, consider Jespersen & Carrara’s [
Artifactual functions are typically characterized in terms of intentions. Moreover, as artifacts (especially technical artifacts) are typically characterized in terms of functions, they are typically characterized in terms of intentions. Intentions, functions and artifacts are generally assumed to be inextricably linked with one another.
There may be nonetheless some limitations with the intention-based theory of functions as well as the intention- and function-based theory of artifacts, in particular when it comes to an ontological analysis of the malfunctioning of artifacts that can be found outside the engineering domain. To illustrate this point, let us consider the following two scenarios: • At time t, this pebble (pebble1) is intended to be used to dispel evil spirits. • At time t’, that pebble (pebble2) is intended to be used to keep papers in place (regardless of whether it is actually so used or not).
As one (if not the only) prevailing view goes, pebble1 at time t is an artifact, as it may be thought of as
an amulet, and Baker considers an amulet to be “a paradigm case of a technical artefact” [14, p. 84]. By
contrast, pebble2 at time t’ is not an artifact and it is rather a natural object to be intended to be used
for some purpose — notwithstanding Borgo & Vieu’s [15] noteworthy exceptional view that, at time t’,
a new material entity (say paperweight2) constituted by pebble2 comes into being and paperweight2
is an artifact (see also [16] and [
Now according to one version of the intention- and function-based theory of artifacts, artifacts are ontologically explicable in terms of “design functions”, rather than in terms of “use functions”. Roughly, design functions are the functions imposed on artifact by their designer(s), and use functions are what users actually use artifacts for, as when a screwdriver is used to open a paint can (see, e.g., [17, 18, 19] for more on design and use functions). Consequently, whether an entity is an artifact or not is determined by whether the entity is intended by the designer, or not (e.g., by the user).
It may be nevertheless contentious whether this version of the intention- and function-based theory
of artifacts can well accommodate the prevailing view (currently under consideration) that pebble1 at
time t is an artifact but pebble2 at time t’ is not. This is because it is challenging to articulate clearly
the ontological diference between the intention to use pebble1 to dispel evil spirits and the intention
to use pebble2 to keep papers in place. Furthermore, given that the term “design” refers to a planning
process for production in the engineering domain, it would be all the less clear to think that pebble1 at
time t is intended by the designer but pebble2 at time t’ is by the potential user, because using pebble1
to dispel evil spirits is nomologically impossible in light of (our knowledge of) the laws of nature in our
actual world, but using pebble2 to keep papers in place is nomologically possible (cf. [
All these considerations can motivate exploring artifactual functions and artifacts diferently from
the intention-based theory of functions as well as the intention- and function-based theory of artifacts.
One possible alternative approach is a realizable-centered approach to artifacts [
The realizable-centered approach is naturally combined with the supposition that (artifactual) functions
are a kind of realizable entities [
There is however a long-standing debate over the status of functions within the BFO category of
realizable entity. According to the latest, ISO-standard version of BFO (ISO/IEC 21838-2), Function is a
subtype of BFO:Disposition [
It has been objected however that Function should not be a subtype of BFO:Disposition but rather a subtype of BFO:Realizable entity, as distinct from both BFO:Disposition and BFO:Role [17, 16]. Note that a role in BFO is a realizable entity that (1) exists because the bearer is in some special physical, social, or institutional set of circumstances in which the bearer does not have to be (optionality), and (2) is not such that, if this realizable entity ceases to exist, then the physical make-up of the bearer is thereby changed (external grounding) — examples being the role of being a student and the role by a stone of marking a boundary. There is also the opinion that BFO could embrace a pluralist conception of functions: some functions are BFO:dispositions and other functions are BFO:realizable entities that are neither BFO:dispositions nor BFO:roles (and/or other functions are BFO:roles) [19, 20].
Quite importantly, this debate concerns whether the BFO dispositional theory of functions (i.e.,
BFO:Function) can account for the phenomenon of malfunctioning, because malfunctioning may be well
analyzed in terms of the lack of a corresponding BFO:disposition and this conception of malfuctioning
would imply that functions exist without relevant BFO:dispositions, hence the categorical independence
of Function from BFO:Disposition. According to Jespersen & Carrara’s [
For another example, Jansen [16, p. 124] distinguishes between “having a malfunction” and “being malfunctioning” and characterize them as follows: • A material object x has a malfunction at t with respect to the function to F-in-situation-S, if and only if x has the function to F-in-situation-S but would not F in S, because, at t, it does not have the disposition to F-in-situation-S. • A material object x is malfunctioning at t with respect to the function to F-in-a-situation-S, if and only if, at t, x is in situation S but does not F.
On the background of these definitions, a car with a fuel-soaked ignition plug has a malfunction even if nobody wants to drive it, as it would not start when one would try to do so. It is actually malfunctioning if one tries to start it but does not accomplish to do so because of the fuel-soaked ignition plug. Based on this analysis of the notions of malfunction and malfunctioning, BFO:Function would make malfunction(ing) conceptually impossible. For if all functions to do something are identical with the disposition to do so, the car would loose its function to drive once it looses its disposition. If malfunctioning requires the presence of a function and the absence of the respective disposition, this can, according to the BFO account, never be the case, as the absence of the disposition would imply that the the function is absent as well. And it is no way out to say that the oil-drop only masked the disposition of the car because there was in fact an intrinsic physical change that destroyed the disposition. If the car is chained to a massive piece of concrete, its intrinsic properties remain the same, and hence also, according to BFO, its dispositions. But no car has, of course, the function to-drive-in-spite-of-being-chained-to-massive-concrete-blocks.
To generalize, there are two types of realizable-based theories of artifactual functions which we will call the “disposition-grounded” and “disposition-free”. On the one hand, both theories of artifactual functions share the view that Function is a subtype of Realizable entity. On the other hand, the dispositiongrounded theory (as illustrated by BFO:Function) says that functions depend (existentially) on some relevant disposition(s), or that functions are even a subtype of dispositions, whereas the disposition-free theory counters that functions are (existentially) independent of dispositions. In the following, we will assume the disposition-grounded theory in developing a realizable-centered account of malfunctioning artifacts. For one thing, at least some (if not all) ideas underlying the disposition-free theory may be captured within the framework of the disposition-grounded theory (combined with a realizable-centered approach to artifacts). We will illustrate this point in Section 3.
We will present some basic notions used in a realizable-centered approach to artifacts [
An extrinsic disposition is a realizable entity that is borne, at least partially, in virtue of the way the
world that is external to the bearer is (see [
While it is a standard research question in the ontology of artifacts what is the correct (and ideally single) definition of the existing term “artifact”, this approach adopts the diferent methodology of investigating the two newly introduced notions, canonical artifact and usefact, which help to understand very diferent usages of the term “artifact” in diferent domains.
A canonical artifact is something that is intentionally produced for some purpose. The notion of
canonical artifact corresponds to the traditional account of artifacts in philosophy of artifacts (e.g.,
[28]), hence the term “canonical artifact”. For illustration, when I intentionally shaped and dried a
certain amount of clay in such a way as to be able to contain liquid, the resulting clay pot is a canonical
artifact because it is intentionally produced for the purpose of being able to contain liquid. As we will
discuss in Section 4, pebble1 at time t can be seen as a kind of canonical artifact whose intentional and
“production” dimensions merit special treatment (see [
In contrast, a usefact is something that is intended to be used for some “accidental purpose”, which means (without recourse to the complex distinction between essentiality and accidentality) some purpose that is diferent from the purpose (if any) for which the entity was intentionally produced. In particular, a natural object is a usefact when it is intended to be used for any purpose, because it was not intentionally produced for any purposes (or this is an assumption that we make in understanding natural objects). To take an example in Section 2.1.2, pebble2 is a usefact at time t’ because it is intended to be used for the accidental purpose of keeping papers in place, or particularly, for the trivial reason that it is a natural object that is intended to be used for some purpose. Concerning the usage of the notion of usefact, Borgo & Vieu’s [15] formal theory of artifacts is a case in point, as their term “artifact” can be construed as referring to usefacts as well as canonical artifacts (see Section 2.1.2).
After explaining the basic idea of the malfunctioning of artifacts (Section 3.1), we will investigate, by analyzing examples based on the ontology of realizable entities, the malfunctioning of material artifacts when the term “material artifact” refers to material canonical artifacts (Section 3.2) and material usefacts (Section 3.3). To recall, we will assume a theory of artifactual functions that is particular-oriented, realizable-centered and disposition-grounded (see Section 2.1).
There are so many accounts of malfunctioning in the literature that it is dificult to give even a general overview (see, e.g., [21, 22] for a general survey). Following e.g., Artiga [20], we will presuppose the following general characterization of the malfunctioning of artifacts:
Proposition 1 (the malfunctioning of material artifacts) A material artifact is malfunctioning if and only if (i) it has some function and (ii) it does not perform this function in the way it is supposed to.
Some clarifications on Proposition 1 are in order. Firstly, Proposition 1 is about the malfunctioning of material artifacts. By focusing on this proposition, we will leave aside the potentially highly complex issue of the malfunctioning of abstract artifacts, including that of softwares [ 23] and fictional characters [24]. Moreover, we are mainly interested in the malfunctioning of particular material artifacts, as we assume a particular-oriented theory of artifactual functions (see Section 2.1.1).
Secondly, terms such as “(material) artifact”, “have a function” and “perform a function” figuring in Proposition 1 are intended to be neutral as to which theories of artifacts and/or functions are adopted. In what follows we will interpret this proposition based on a realizable-centered and disposition-grounded theory of artifactual function, according to which functions are a kind of realizable entities that are grounded in some relevant disposition(s) (see Sections 2.1.2 and 2.1.3).
Thirdly, we will use, whenever possible, the verb phrase “to be malfunctioning” in discussing the malfunctioning of artifacts. For one thing, we want to remain agnostic as to the meanings of other malfunctioning-related terms, including “failure” and “fault” [25, 26] as well as “dysfunction” [27], which are frequently used in engineering and in biology, respectively. For another, we avoid using the noun “malfunction” and the simple tense of the verb “to malfunction” because they could yield terminological confusion in view of the distinction between “malfunction” and “malfunctioning” (as advocated by Jansen [16]; see Section 2.1.3).
Fourthly and finally, the malfunctioning of artifacts has some normative dimension, as is shown by item (ii) of Proposition 1 (“in the way it is supposed to”) and is explicitly highlighted by Baker [14]. At the same time, Proposition 1 is neutral as to the source of the normativity of the malfunctioning of artifacts. In the ensuing we will analyze the normativity in question by utilizing a realizable-centered approach to artifacts, to wit, in terms of a special kind of realizable entity (“intentional realizable entity”; see Section 3.2) involving a relevant intentional dimension.
For sake of simplicity, the following discussion will presuppose the “continuity view” of diachronic
identity with respect to production (see [
Example 1 (material canonical artifact) At time t1, a particular amount of clay (clay1) was shaped and dried to be able to contain liquid.
We will focus on the intrinsic disposition d1 to contain liquid that clay1 comes to bear at time t1 purely
in virtue of its specific physical (e.g., container-like) structure. According to Toyoshima et al. [
Definition 2 (novel realizable entity) novel realizable entity =def. A realizable entity r such that the bearer has no realizable entity r’ such that (i) r’ exists before r came into being and (ii) if r is realized in a process, then r’ is realized in the same process.
First, d1 is an intentional realizable entity, as it comes into being the intentional act of shaping and drying clay1 and its realization (i.e., a process of clay1 containing liquid) is determined by the goal that clay1 should contain liquid.1 Second, d1 is a novel realizable entity, as there is no pre-existing realizable entity r’ of clay1 such that, if d1 is realized in a process of clay1 containing liquid, then r’ is also realized in this process.
Toyoshima et al. [
Definition 3 (novel intentional realizable entity) novel intentional realizable entity =def. An intentional realizable entity that is a novel realizable entity.
Definition 4 (material canonical artifact) material canonical artifact =def. A material entity that bears a novel intentional realizable entity.
According to these definitions, d1 is a novel intentional realizable entity and clay1 is a material canonical artifact at time t1.
Let us now investigate the malfunctioning of material canonical artifacts by considering the following example:
Example 2 (the malfunctioning of material canonical artifacts) At time t2 (later than t1), clay1 is intended to be used to, but fails to, contain a particular amount of liquid (liquid2).
For expository purposes, we will first consider the normative dimension of the malfunctioning of clay1 at time t2 (item (ii) of Proposition 1). Arguably, the normativity under question stems from the fact that clay1 is intended to be used to contain liquid2 at time t2, or more specifically, from the intention to use clay1 to contain liquid2 at time t2 ([29]; see Section 4). One realizable-based way of interpreting the claim is to think that, at time t2, clay1 comes to bear an extrinsic disposition d2 to contain liquid2 such that d2 exists in virtue of the existence of the intention to use clay1 for that purpose at time t2. Note that d2 is an intentional realizable entity, as it comes into being through the intentional act using clay1 to contain liquid and its realization is determined by the goal of clay1 containing liquid. We propose to understand the normativity of the malfunctioning of clay1 at time t2 in terms of the intentional realizable entity d2.
We will next consider a function that is involved in the malfunctioning of clay1 at time t2 (item (i)
of Proposition 1) or more specifically which realizable entity of clay1 is a relevant function at time t2
(according to a realizable-centered theory of artifactual functions). There are at least two realizable
entities that can be a function relating to the malfunctioning of clay1 at time t2.
1See [
We highlight that dliquid2 is intimately connected with but diferent from the intrinsic disposition d1, as d1 can be realized in a process of clay1 containing any instance (e.g., liquid2) of the type Liquid, whereas dliquid2 can be realized in a process of clay1 containing liquid2. Moreover, clay1 is a material canonical artifact at time t2 (as well as at time t1) in virtue of bearing dliquid2 because, just as with d1, dliquid2 is a novel intentional realizable entity.
We can thus specify three possible interpretations of Example 1 depending on whether d1 is a function or not and whether dliquid2 is a function or not, while excluding the possibility that neither d1 nor dliquid2 is a function: • Interpretation 1: d1 is a function and dliquid2 is also a function. • Interpretation 2: d1 is a function but dliquid2 is not a function.
• Interpretation 3: d1 is not a function but dliquid2 is a function.
We underscore that Interpretations 1 and 3 may help to understand some ideas behind the
dispositionfree theory of artifactual functions in the present context of the disposition-grounded theory (see
Section 2.1.3). According to these two interpretations, dliquid2 is a function. Since dliquid2 is an
extrinsic disposition, these interpretations allow for the possibility that some extrinsic dispositions can
be functions. To be sure, functions that are extrinsic dispositions may not be disposition-free functions,
as they depend (existentially) on some intrinsic disposition(s) [
We will now analyze the case in which d1 is a function (Interpretations 1 and 2) and the case in
which dliquid2 is a function (Interpretations 1 and 3). For this purpose, we will introduce the relation of
“being non-novel because of” between realizable entities [
Definition 5 (is non-novel because of) A realizable entity r is non-novel because of a realizable entity r’ =def.
There exists some bearer b such that (i) b bears r and (ii) b bears r’ and (iii) r’ exists before r comes into being and (iv) if r is realized in a process, then r’ is realized in the same process.
In the first case, clay1 is malfunctioning with respect to d1, and thus d1 is not realized, at time t2. Now recall that d2 can be helpful in analyzing the normativity of the malfunctioning of clay1 at time t2. We can apply the relation specified by Definition 5 to the analysis of the connection between d1 and d2 in such a way that d2 is non-novel because of d1 (informally: the canonical artifact clay1 is intended to be used for the same purpose for which it was intentionally produced). One consequence of this analysis is that, since d1 is not realized, d2 is not either, according to Definition 5. This analysis can apply, mutatis mutandis, to the second case, where clay1 is malfunctioning with respect to dliquid2. That is to say, dliquid2 is not realized and d2 is non-novel because of dliquid2.
We can generalize these observations as to d1, dliquid2 and d2 into the following realizable-based
characterization of the malfunctioning of material canonical artifacts:
2We leave out of scope the interpretation that, in contrast with dliquid2, d2 has as (part of) its realization a process of the user’s
desire to use clay1 to contain liquid2 coming to be satisfied. See also [
Proposition 2 (the malfunctioning of material canonical artifacts and realizable entities) A material canonical artifact a is malfunctioning if and only if: there exist some function f and some intentional realizable entity r such that (i) a bears f and (ii) a bears r and (iii) f is a novel intentional realizable entity and (iv) f is not realized and (v) r is non-novel because of f.
Informally speaking, Proposition 2 says that a material canonical artifact a being malfunctioning amounts to the thesis that a is intended to be used for the same purpose for which it was intentionally produced (ii, iii, v), but a fails to perform the function f in the way it is intended to be (i, iv).
Let us now turn to investigating the malfunctioning of material usefacts (under the continuity view of use). We will introduce the following example of material usefacts that is parallel to the motivating example of pebble2 at time t’ (introduced in Section 2.1.2):
Example 3 (material usefact) At time t3, a particular pebble (pebble3) is intended to be used to keep papers in place (regardless of whether it is actually so used or not).
We will focus on two realizable entities involved in this example. One is the extrinsic disposition d3 to keep papers in place, which pebble3 comes to bear at time t3 partly in virtue of the existence of the intention to use pebble3 for that purpose. As with d2 (borne by clay1 at at time t2) in Example 2, d3 is an intentional realizable entity.
The other is the intrinsic disposition d3* to keep papers in place, which pebble3 has borne since before time t3 purely in virtue of its specific physical (e.g., solid) structure. As compared to d1 (borne by clay1 at times t1 and t2) in Examples 1 and 2, d3* is not an intentional realizable entity and, a fortiori, it is not a novel intentional realizable entity, as pebble3 is a natural object and it was not intentionally produced for any purposes — notably, for the purpose of keeping papers in place.
The term “material usefact” is defined in terms of the notions of intentional realizable entity and
novel intentional realizable entity (Definitions 1 and 3) as well as the relation of “being non-novel
because of” (Definition 5) as follows [
Definition 6 (material usefact)
material usefact =def. A material entity that bears an intentional realizable entity which is
non-novel because of some realizable entity that is not a novel intentional realizable entity.
According to this definition, pebble3 is a material usefact at time t3 because d3 is an intentional
realizable entity, d3* is not a novel intentional realizable entity and d3 is non-novel because of d3*.
Moreover, we can say that d3 is a “usefactual realizable entity”, this term being defined as follows
[
Definition 7 (usefactual realizable entity) usefactual realizable entity =def. An intentional realizable entity which is non-novel because of some realizable entity that is not a novel intentional realizable entity.
We will now analyze the malfunctioning of material usefacts by considering the following example: Example 4 (the malfunctioning of material usefacts) At time t4, pebble3 is intended to be used to, but fails to, keep a paper in place (e.g., owing to a very strong gust of wind).
To consider the malfunctioning of the material usefact pebble3 at time t4, we will focus on the extrinsic disposition d4 to keep a paper in place that pebble3 comes to bear at time t4 partly in virtue of the existence of the intention to use pebble3 for that purpose at time t4. Regarding the normativity of the malfunctioning of pebble3 at time t4, we can understand it in terms of d4, as it is a usefactual realizable entity — in particular, d4 is non-novel because of d3*. Moreover, it is plausible to think that d4 is a function that is involved in the malfunctioning of pebble3 at time t4.
We can generalize this observation about d4 into the following realizable-based characterization of the malfunctioning of material usefacts:
Proposition 3 (the malfunctioning of material usefacts and realizable entities) A material usefact a is malfunctioning if and only if there exist some function f such that (i) a bears f and (ii) f is a usefactual realizable entity and (iii) f is not realized.
To put it informally, Proposition 3 says that a material usefact a being malfunctioning amounts to the thesis that a is intended to be used for some accidental purpose (ii), but a fails to perform the function f in the way it is intended to be (i, iii).
One important consequence of Proposition 3 is that no material artifact can be malfunctioning if the
term “function” therein refers to a usefactual realizable entity that is being realized. To illustrate this
point, we will introduce the term “usefactual realized entity” that is defined as follows [
The notion of usefactual realized entity can be useful in analyzing existing conceptions of functions that are based exclusively on actual use (such as “sporadic user-intended function” [18] and “narrow use function” [19]) and, according to Proposition 3 and Definition 8, no usefactual realized entity can be a function that is involved in the malfunctioning of any material usefacts. At the same time, it warrants exploration whether and how Proposition 3 could be extended to cases where, for example, a pebble is so conventionally used to keep papers in place that it might have a use function (to wit, “conventional user-intended function” [18]) to keep papers in place even when there is no paper on my desk and it is not currently intended for that use purpose.
In this section we will discuss some selected works on malfunctioning (of artifacts) in order to elucidate our realizable-centered account of malfunctioning artifacts and to specify possible directions of research in which to further develop our approach.
First of all, our realizable-centered account of malfunctioning artifacts is akin to Roberts’s [29] Dispositional Account of technical Functions (DAF), some of whose core ideas can be presented as follows:
DAF: If x is a technical function of artifact A, then x is ultimately referring to a disposition of A. [29, p. 7; note that the term “ultimately referring to” therein can be arguably omitted.] DAF Malfunction: Artifact A malfunctions only if some appropriate use case R is carried out and the expected disposition does not manifest. [29, p. 11] It is straightforward to see that (DAF) and (DAF Malfunction) are comparable to our realizable-centered account of artifactual functions and malfunctioning artifacts, as is witnessed by the intimate connection between (DAF Malfunction) and Proposition 2 (see Section 3.2) — in particular, items (iv) and (v) thereof.
At the same time, there are some important diferences between Roberts’s and our accounts of the malfunctioning of artifacts. First, the term “artifact” refers to canonical artifacts for him, whereas it can refer to usefacts as well for us (Proposition 3; see Section 3.3). Second, he seems to use the term “disposition” to refer (primarily) to an intrinsic disposition, whereas we leave open the possibility of the existence of functions that are extrinsic dispositions, as it may serve as surrogates for disposition-free functions (see Section 3.2).
Third, (DAF Malfunction) implies that “phantom functions are a severe case of malfunctioning” [29, p. 11], where the term “phantom function” was proposed by Preston [30] to refer to a function of a kind of artifact that the artifact is “constitutionally incapable of performing” (ibid., p. 177; see also Krohs’ [27] notion of “type-malfunction”). We can illustrate the notion of phantom function with the motivating example of pebble1 at time t (see Section 2.1.2): pebble1 has a phantom function to dispel evil spirits at time t.
Pace Roberts, we find it problematic to integrate the notion of phantom function into a realizablecentered theory of artifactual functions because it would be committed to the highly controversial thesis that there are some realizable entities whose realizations are nomologically impossible (but see [31] for a dissenting voice). A diferent interpretation of the pebble1 at time t scenario can be derived from Baker’s [14] account of malfunctioning artifacts (see Section 2.1.1). According to this interpretation, pebble1 at time t has the intended function to dispel evil spirits, but it cannot malfunction because it is physically impossible to perform this intended function.
In contrast with Roberts’s and Baker’s function-based accounts of artifacts, our realizable-centered account of artifacts and their malfunctioning can ofer a third view of pebble1 at time t by deploying Koslicki & Massin’s [32] analysis of “faith-based artifacts” in terms of what they call “placebo capacities”. According to their explanation, a faith-based artifact is an artifact of a kind that is intended and believed by its creators and users to perform a function which it does not in fact perform. Examples of faith-based artifacts include amulets such as pebble1 at time t. They suggest that faith-based artifacts can be explicable in terms of the notion of placebo capacity: a capacity to subjectively satisfy an agent’s desire to produce the relevant efect in the presence of a belief by the agent that the entity is able to bring about this efect. For instance, amulets can be ascribed the placebo capacities to subjectively satisfy the user’s desire to dispel evil spirits in the presence of the relevant belief on the part of the agent that they can in fact dispel evil spirits. But since amulets in fact lack the capacity to dispel evil spirits, the user’s desire to dispel evil spirits can be only subjectively, but not objectively, satisfied.
The notion of placebo capacity enables us to think alternatively that pebble1 at time t may not have
a function to dispel evil spirits — which is a phantom function — and therefore it cannot malfunction
(by Proposition 1 or 3); but it still has the placebo capacity to subjectively satisfy the user’s desire to
dispel evil spirits and this capacity is a realizable entity whose realizations are nomologically possible.
Moreover, if a placebo capacity can be generally characterized as a novel intentional realizable entity
(to wit, a novel intentional extrinsic disposition), we can say that pebble1 at time t is a canonical
material artifact in virtue of bearing the placebo capacity under consideration, in alignment with a
realizable-centered account of malfunctioning artifacts (see again [
Finally, we will discuss Bahr’s [18, p. 96] “malfunctioning desideratum” for a satisfactory account of artifactual functionality, accommodating the functionality of both technical artifacts and artworks. This desideratum consists of three more specific criteria. The first specific criterion is as follows: (1) An account A should allow the ascription of the function of -ing to a technical artifact, an art work or an instance of an art work a although a fulfills that function inefectively only if there are good reasons for the assumption that a still fulfills the function of -ing. To illustrate criterion (1) with Bahr’s example: “If the blade of a knife is blunt, it might not be able to fulfill the function of cutting optimally, although one can still cut something with this knife. Although the knife might fulfill its function rather badly, we would nonetheless like to call cutting one of its functions” [18, p. 91].
This criterion for malfunctioning can be interpreted in terms of Proposition 2, more specifically as
the case where there is some problem with the “categorical basis” of the relevant function (which is a
realizable entity) (see [
The second specific criterion of the malfunctioning desideratum is as follows [ 18, p. 96]: (2) An account A should allow the ascription of the function of -ing to a technical artifact, an art work or an instance of an art work a although a currently does not fulfill the function due to inefective handling only if (a) there are good reasons for the assumption that a was intentionally created for -ing or a itself and/or tokens of the type a belongs to has/have regularly been used to perform the function of -ing and (b) there are good reasons for the assumption that a can potentially perform the function of -ing.
To illustrate criterion (2) with Bahr’s example: “if I fail to push the right button to turn on my electric kettle or if I forget to plug it in or to fill it with water, the kettle will not fulfill its function of boiling water. Nevertheless […] it is a function of the kettle to boil water, because, if the kettle had been handled correctly, it would have fulfilled its function of boiling water” [ 18, p. 91]. In addition: “If someone tries to open a beer bottle with a lighter and fails (which supposedly happens quite often […]), […] this is even a function of this particular lighter, provided that it can potentially open beer” (ibid., p. 94).
This criterion for malfunctioning can be partially interpreted in terms of Proposition 2, specifically as the case where the relevant function (which is a realizable entity) is not realized because it is not “triggered” — the cases of malfunctioning relevant to “regular use” being reserved for future work in connection with Proposition 3 (see Section 3.3). To illustrate this with Example 2, d1 is not realized at time t2 because the handle of clay1 is improperly held.
The third specific criterion of the malfunctioning desideratum is as follows [ 18, p. 96]: (3) An account A should allow the ascription of the function of -ing to a technical artifact, an art work or an instance of an art work a although a is currently not capable of -ing only if (a) there are good reasons for the assumption that a was intentionally created for -ing or a itself and/or tokens of the type a belongs to has/have regularly been used to perform the function of -ing and (b) there are good reasons for the assumption that a can be fixed to win back its capability of -ing.
To illustrate criterion (3) with Bahr’s example: “If my cofee machine is broken, but not unfixable, I still want to say that it is a function of the cofee machine to make cofee” [ 18, p. 95]. Furthermore: “If one of my chairs has a broken leg and cannot serve as a ladder, […] I can still ascribe the function of serving as a ladder to that chair as long as I believe that the chair can be fixed and regain its capability to serve this purpose, because it is such a regular thing to use chairs for that purpose that this function is strongly connected to these artifacts” (ibid., pp. 95-96).
Incorporating this criterion for malfunctioning into our realizable-centered account will necessitate careful consideration of the notion of what which we may call “repairability”, and what is expressed by Bahr in terms like “can be fixed to win back [synonym: ‘regain’] its capability” and being “not unfixable”. While scrutinizing the notion of reparability lies outside the purview of this paper (see, e.g., [33]), we remark that, from a realizable-centered perspective, repairability would be characterized, at ifrst approximation, as an intrinsic disposition to regain some relevant realizable entity (see [ 34] for a pointer).
We outlined an ontological unifying approach to malfunctioning artifacts that is theoretically undergirded by the ontology of realizable entities. For this purpose, we drawed upon a realizable-centered approach to artifacts and provided an ontological analysis of the malfunctioning of material canonical artifacts and material usefacts. The key idea is that an artifact is malfunctioning when some relevant intentional realizable entity thereof is not realized. By discussing related work, we elucidated our realizable-centered account of the malfunctioning of artifacts and specified some future lines of inquiry, such as an ontological analysis of regular/conventional use and repairability within our framework.
For Fumiaki Toyoshima: This work is part of the project “An Ontology of Production, Products, and By-Products” (2023-2027; grant number #212493) funded by the Swiss National Science Foundation (SNSF). FT is financially supported by SNSF. FT’s participation in the relevant workshop is made possible by funding from the Swiss Academy of Humanities and Social Sciences (SAHSS). The research has been conducted in cooperation with the Swiss Center for Ontological Research (SCOR).
For Ludger Jansen: This paper has been started as part of the project “Learning from Nature”, funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation, project 492191929) and ifnished during a DFG-funded research stay at the Centre for Advanced Study “Access to cultural goods in digital change: art historical, curatorial, and ethical aspects” (KFG 33; project 449836922).
During the preparation of this work, the authors used ChatGPT in order to: Grammar and spelling check, Paraphrase and reword. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the publication’s content. [13] B. Jespersen, M. Carrara, A new logic of technical malfunction, Studia Logica, 101(3), 547-581, 2013. [14] L. R. Baker, The metaphysics of malfunction, Techné: Research in Philosophy and Technology, 13(2), 82-92, 2009. [15] S. Borgo, L. Vieu, Artifacts in formal ontology, in A. Meijers (Ed.), Handbook of the Philosophy of Science. Handbook of Philosophy of Technology and Engineering Sciences, vol. 9, pp. 273-307, Elsevier, 2009. [16] L. Jansen. Functions, malfunctioning, and negative causation, in A. Christian, D. Hommen, N.
Retzlaf, G. Schurz (Eds.), Philosophy of Science. European Studies in Philosophy of Science, vol. 9, Cham: Springer, pp. 117-135, 2018. [17] J. Röhl, L. Jansen, Why functions are not special dispositions: An improved classification of realizables for top-level ontologies, Journal of Biomedical Semantics, 5, 27, 2014. [18] A. Bahr, What the Mona Lisa and a screwdriver have in common, Grazer Philosophische Studien, 96(1), 81-104, 2019. [19] F. Toyoshima, A. Barton, J.-F. Ethier, Investigating functions in BFO from the viewpoint of extrinsic dispositions. in T. P. Sales et al. (Eds.), Proceedings of the 8th Joint Ontology Workshops (JOWO 2022), Jönköping, Sweden, August 15-19, 2022. CEUR Workshop Proceedings, Vol. 3249, pp. 1-7, 2022. [20] M. Artiga, New perspectives on artifactual and biological functions, Applied Ontology, 11(2), 89102, 2016. [21] F. Compagno, S. Borgo. Ontological analysis of malfunctions: Some formal considerations, in C.
Trojahn, D. Porello, P. P. F. Barcelos (Eds.), Proceedings of the 14th International Conference on Formal Ontology in Information Systems (FOIS 2024), online, July 8-9 and Enschede, Netherlands, July 15-19, 2024. Amsterdam: IOS Press, pp. 149-162, 2024. [22] F. Compagno, Modelling of functions and malfunctions in industrial processes: An applied ontology approach, PhD thesis, University of Trento, Italy, 2025, available online at: https://hdl.handle.net/11572/450850. (Last access on June 14, 2025.) [23] L. Floridi, N. Fresco, G. Primiero, On malfunctioning software, Synthese, 192, 1199-1220, 2015. [24] E. Terrone, Taking abstract artifacts seriously — The functioning and malfunctioning of fictional characters, Philosophies, 8(6), 105, 2023. [25] L. Del Frate, M. Franssen, P. Vermaas, Towards a trans-disciplinary concept of failure for integrated product development, International Journal of Product Development, 14(1-4), 72–95, 2011. [26] L. Del Frate, Preliminaries to a formal ontology of failure of engineering artifacts, in M. Donnelly, G. Guizzardi (Eds.), Proceedings of the 7th International Conference on Formal Ontology in Information Systems (FOIS 2012), Graz, Austria, July 24-27, 2012. Amsterdam: IOS Press, pp. 117-130, 2012. [27] U. Krohs, Dys-, mal-, and non-: The other side of functionality, in J. Gayon, A. de Ricqlès, A. C.
Dussault (Eds.), Functions: From Organisms to Artefacts. History, Philosophy and Theory of the Life Sciences, vol. 32. Springer, Cham, pp. 303-312, 2023. [28] R. Hilpinen, Artifacts and works of art, Theoria, 58(1), 58-82, 1992. [29] M. A. Roberts. A dispositional account of technical functions, Synthese, 204, 108, 2024. [30] B. Preston, A Philosophy of Material Culture: Action, Function, and Mind, New York: Routledge, 2013. [31] C. S. Jenkins, D. Nolan, Disposition impossible, Noûs, 46(4), 732-753, 2012. [32] K. Koslicki, O. Massin, Artifact-functions: A capacity-based approach, in M. J. García-Encinas, F.
Martínez-Manrique (Eds.), Special Objects: Social, Fictional, Modal, and Non-Existent, Springer, pp. 31-52, 2025. [33] T. M. Carvalho, There, I fixed it! On the status and meaning of repair, in M. T. Young, M. Coeckelbergh (Eds.), Maintenance and Philosophy of Technology: Keeping Things Going, New York: Routledge, pp. 86-112, 2024. [34] K. Koslicki, Artifacts and the limits of agentive authority, in M. Garcia-Godinez (Ed.), Thomasson on Ontology, pp. 209-241, Springer Verlag, 2023.