70 Pieter E. Vermaas
relative to context, and then construct a theory by which technical functions are also rela-
tions that artifacts have relative to context. I acknowledge that these latter relations are
still subjective in an ontological sense, but defend that they are objective in an epistemic
sense. By thus minimizing the differences between biological and technical functions,
prospects for a uniform function theory improve, which I explore at the end of this
contribution.
The technical function theory that I construct is drawn from the ICE-function theory
(Houkes and Vermaas 2004; Vermaas and Houkes 2006), in which technical functions are
relations that agents ascribe to artifacts relative to mental states. A fi rst assessment of this
constructed theory seems, however, to immediately reveal a snag, since the theory seems
incapable of accommodating the phenomenon of malfunctioning artifacts. I therefore also
introduce in this contribution a new approach toward understanding malfunctioning. This
approach turns the constructed technical function theory into one that can adequately
accommodate malfunctioning; yet it reveals also a new difference between biological and
technical functional descriptions: artifacts can be taken as malfunctioning only if they can
reasonably be repaired, whereas malfunctioning biological items may be irreversibly
malformed.
I introduce in section 5.2 the distinctions between the epistemic and ontological senses
of objectivity and subjectivity, which I adopt from Searle. Then I present in sections 5.3
and 5.4 a strategy to construct theories by which artifacts have technical functions from
theories by which agents ascribe these functions. I apply this strategy to the ICE theory
in section 5.5 to arrive at my theory in which artifacts have their functions as epistemically
objective and ontologically subjective relations relative to the mental states of designers.
The new approach toward understanding malfunctioning is given in section 5.6. I general-
ize the constructed theory to a uniform “ICE-like” function theory in section 5.7, and
indicate its similarities with Cummins’s theory.
5.2 The Subjectivity of Technical Functions
If biological functions are, by the main theories of such functions, to be taken as objective
relations items have relative to context, and if technical functions are subjective relations
that agents ascribe to artifacts relative to mental states of agents, then the differences
between the two consist of two elements: biological functions are objective whereas tech-
nical functions are subjective, and biological functions are relations items have whereas
technical functions are relations that are ascribed by agents to artifacts. These elements
are related. If technical functions are analyzed as relations ascribed by agents, then the
mental states of the ascribing agents seem somehow constitutive to technical functions,
giving these functions a subjective character. Yet in this contribution I consider the two
elements separately, starting in this section with the fi rst.