ArpSmithFunctionBFO

Function, Role, and Disposition in Basic Formal Ontology

Robert Arp

and Barry Smith

National Center for Biomedical Ontology (NCBO) and New York State Center of Excellence in Bioinformatics and Life
Sciences, University at Buffalo,

rarp@buffalo.edu

phismith@buffalo.edu

ABSTRACT
Numerous research groups are now utilizing Basic Formal
Ontology (BFO) as an upper-level framework to assist in the
organization and integration of biomedical information. This
paper provides elucidation of the three BFO categories of
function, role, and disposition, and considers two proposed
sub-categories of artifactual function and biological function.
The motivation is to help advance the coherent treatment of
functions, roles, and dispositions, to help provide the poten-
tial for more detailed classification, and to shed light on
BFO's general structure and use.

1 INTRODUCTION

Many of the members of the Open Biomedical Ontologies
(OBO) Foundry initiative, including the Gene Ontology, the
Foundational Model of Anatomy, the Protein Ontology, and
the Ontology for Biomedical Investigations (http://www.
obofoundry.org/) are utilizing Basic Formal Ontology
(BFO) to assist in the categorization of entities and relation-
ships in their respective domains of research.

Fig. 1. The continuant categories of BFO.

BFO:entity

continuant

independent continuant

object

object boundary

object aggregate

fiat object part

site

dependent continuant

generically dependent continuant

specifically dependent continuant

quality

realizable entity

function

role

disposition

spatial region

zero-dimensional region

one-dimensional region

two-dimensional region

three-dimensional region

To whom correspondence should be addressed.

Many individuals and groups involved in organizations

such as BioPAX, Science Commons, Ontology Works, As-
traZeneca, and the Computer Task Group utilize BFO as
well.

Fig. 2. The occurrent categories of BFO.

BFO:entity

occurrent

processual entity

process

process boundary

process aggregate

fiat process part

processual context

spatiotemporal region

scattered spatiotemporal region

connected spatiotemporal region

spatiotemporal instant

spatiotemporal interval

temporal region

scattered temporal region

connected temporal region

temporal instant

temporal interval

Versions of BFO in OBO, OWL and first-order logic formats are

maintained by Holger Stenzhorn at http://www.ifomis.org/bfo.
Definitions and other content taken from there have been mod-
ified to provide additional clarity of exposition.

BFO is an upper-level ontology developed to support in-

tegration of data obtained through scientific research. It is
deliberately designed to be very small, in order that is
should be able to represent in consistent fashion the upper-
level categories common to domain ontologies developed by
scientists in different domains and at different levels of gra-
nularity. BFO adopts a view of reality as comprising (1)
continuants, entities that continue or persist through time,
such as objects, qualities, and functions, and (2) occurrents,
the events or happenings in which continuants participate.
The subtypes of continuant and occurrent represented in
BFO are presented in Figures 1 and 2 (Grenon and Smith,
2004; Smith and Grenon, 2004; http://www.ifomis.uni-
saarland.de/bfo/).

Arp & Smith

2 FUNCTION, ROLE, AND DISPOSITION

Use of the term `function' is common in descriptions of
molecular and cellular processes, as in assertions such as:

the function of the kidney of Mus musculus is to filter
out waste and water which become urine,

Arabidopsis thaliana has a multifunctional protein

there are several folD bifunctional proteins in Campy-
lobacter jejuni.

Functions thus play a central role in the Gene Ontology
(http://www.geneontology.org/).

What, however, of the non-biological functions of arti-

facts such as screwdrivers, microplates, or pycnometers?
Are there both designed (artifactual) and natural (biological)
functions, representing distinct subtypes of the more general
category of BFO:function?

A related issue is that of the use of the terms `function'

and `role'. These are distinguished by BFO as representing
two distinct categories (Figure 1), but outside BFO circles
they are often used interchangeably, as when function is
defined as `the role that a structure plays in the processes of
a living thing'. Analogous difficulties arise with regard to
the terms `disposition' and `tendency', as in: `blood has the
tendency or disposition to coagulate', `a hemophiliac has the
disposition or tendency to bleed an abnormally large amount
of blood', and `that patient has suicidal dispositions or ten-
dencies'.

In this paper, we attempt to elucidate the categories of

function, role, and disposition in BFO. We also describe two
sub-type categories of function, the artifactual and the bio-
logical, and provide definitions for each.

Within the context of BFO, one should correctly state:

the (or a) function of the heart is to pump blood

the role of the surrogate is to stand in for the patient

blood has the disposition to coagulate

that patient has suicidal tendencies

To see why this is so, we need first to consider BFO's more
general approach to classification.

In BFO, all entities are divided into continuants and oc-

currents; continuants in turn are divided into independent
and dependent. Independent continuants are things (the ob-
jects we see around us every day) in which dependent conti-
nuants--such as qualities, functions, roles, dispositions--
can inhere.

Dependent continuants stand to their bearers in the rela-

tion of existential dependence: in order for them to exist,
some other (independent) entity must exist. For example,
instances of qualities such as round and red are dependent
continuants in that they cannot exist without being qualities
of some independent continuant such as a ball or a clown's
nose. So too, functions, roles, and dispositions exist only
insofar as they are functions, roles, and dispositions of some
(one or more) independent continuant. The function of my

heart is an instance of the BFO type function, and so also is
the function of your heart.

One major subcategory of dependent continuants in BFO

is that of realizable entity. Realizable entities are defined by
the fact that they can be realized (manifested, actualized,
executed) in occurrents of corresponding sorts. Examples of
realizable entity types include: the function of the liver to
store glycogen, the role of being a doctor, the disposition of
metal to conduct electricity.

Realizable entities are entities of a type whose instances

are typically such that in the course of their existence they
contain periods of actualization, when they are manifested
through processes in which their bearers participate. They
may also exhibit periods of dormancy where they exist by
inhering in their bearers, but are not manifested, as for ex-
ample, in the case of certain diseases. Some realizables,
such as the function of a sperm to penetrate an ovum, may
be such that they can be manifested only once in their life-
time; or, as again in the case of sperm, they are realized only
in very rare cases.

We are now in a position where we can define function,

role, and disposition.

2.1

Function

A function f is

(1)

a realizable dependent continuant.

Thus,

(2)

it has a bearer, which is an independent continuant,
and

(3)

it is of a type instances of which typically have realiza-
tions; each realization is

a process in which the bearer is participant

that occurs in virtue of the bearer's physical make-
up,

and this physical make-up in something which that
bearer possesses because of how it came into be-
ing.

Examples include: the function of a birth canal to enable
transport and the function of a hammer to drive in nails. The
process under a. may be specified further as an end-directed
activity, by which we mean in the biological case something
like: an activity that helps to realize the characteristic physi-
ology and life pattern for an organism of the relevant type.
Each function has a bearer with a physical structure which,
in the biological case, the bearer has naturally evolved to
have (as in a hypothalamus secreting hormones) or, in the
artifact case, the bearer has been constructed to have (as in
an Erlenmeyer flask designed to hold liquid) (Ariew and
Perlman, 2002).

It is not accidental or arbitrary that a given eye has the

function to see or that a given screwdriver have been de-
signed and constructed with the function: to fasten screws.
Rather, these functions are integral to these entities in virtue

Function, Role, and Disposition in Basic Formal Ontology

of the fact that the latter have evolved or been constructed to
have a corresponding physical structure.

If a continuant has a function, then it is built to exercise

this function reliably on the basis of this physical structure.
But again: a function is not in every case exercised or mani-
fested. Its bearer may be broken; it may never be in the right
kind of context. Hence, when we say that a given structure
is designed in such a way as to bring about a certain end
reliably, then this reliability presupposes the fulfillment of
certain conditions, for example of an environmental sort.

On the level of instances, this can be stated as: if f is the

function of c, then (in normal circumstances), c exercises f.

On the level of universals, as: if F is the function univer-

sal exemplified by instances of the independent continuant
universal C, then (in normal circumstances) instances of C
participate in process instances which are realizations of F.
The implications of this analysis for the treatment of func-
tions in the Gene Ontology are outlined in Hill, Smith,
McAndrews-Hill, and Blake (2008).

2.2

Role

In contrast to function, role is a realizable entity whose ma-
nifestation brings about some result or end that is not typical
of its bearer in virtue of the latter's physical structure. Ra-
ther, the role is played by an instance of the corresponding
kind of continuant entity because this entity is in some spe-
cial natural, social, or institutional set of circumstances
(http://www.ifomis.org/bfo).

Examples include: the role of a chemical compound to

serve as analyte in an experiment, the role of penicillin in
the treatment of a disease, the role of bacteria in causing
infection, the role of a person as student or surgeon.

What is crucial for understanding a role--as distinct from

a function--is that it is a realizable entity that an indepen-
dent continuant can take on, but that it is not a reflection of
the in-built physical structure of that independent conti-
nuant. Certain strains of Escherichia coli bacteria have the
role of pathogen when introduced into the gut of an animal,
but they do not have this role when merely floating around
in a pool of water. A heart has the function of pumping
blood; but in certain circumstances that same heart can play
the role of dinner for the lion.

Roles are optional, and they often involve social ascrip-

tion. This is why a person can play the role of being a law-
yer or a surrogate to a patient, but it is not necessary for
persons that they be lawyers or surrogates.

So, when researchers are considering whether some rea-

lizable entity is a function or a role, the question to ask is
this: Is the realizable entity such that its typical manifesta-
tions are based upon its physical structure? If so, then it is a
function. Or, is the realizable entity such that its typical ma-
nifestation is a reflection of surrounding circumstances, es-
pecially those involving social ascription, which are option-
al? If so, then it is a role.

From this perspective, it is incorrect to make assertions

such as:

the role of the heart is to pump blood;

driving nails is a role that this hammer fulfills;

the function of the surrogate is to stand in for the pa-
tient;

the function of James is to serve as my servant.

2.3

Disposition versus Tendency

It is common to find researchers making claims like: `water
has the disposition to rise in a tube', `Carbon-10 has a dis-
position to decay to Boron-10', and `the cell wall is dis-
posed to filter chemicals in endocitosis and exocitosis.' A
disposition is a realizable dependent continuant that typical-
ly causes a specific process in the object in which it inheres
when the object is introduced into certain specific circums-
tances. In addition, these processes occur as a result of the
object's physical structure (Jansen, 2007).

A disposition invariably leads to a certain result given

certain circumstances. Consider: the disposition of a car
windshield to break if struck with a sledgehammer moving
at 100 feet per second; the disposition of a cell to become
diploid following mitosis; the disposition of a magnet to
produce an electrical field.

Contrasted with a disposition is a tendency, which is a

realizable dependent continuant that potentially (not invari-
ably or definitely) causes a specific process in the object in
which it inheres when the object is introduced into certain
specific circumstances as a result of the object's physical
structure property.

Examples include: the tendency on the part of a hemophi-

liac to bleed an abnormally large amounts of blood and the
tendency on the part of a person who smokes two packs of
cigarettes a day throughout adulthood to die of a disease at a
below average age. A patient may have a tendency, and not
a disposition, to commit suicide; while a crystal vase has a
disposition, and not a tendency, to break when it hits the
ground after being dropped from a tall building. We are
referring to tendencies when we refer to genetic and other
risk factors for specific diseases.

3 TWO SUB-CATEGORIES OF FUNCTION

It is possible that BFO has failed to recognize categories or
sub-categories of entities existing in reality. The ontology is,
however, developed on the basis of a principle of scientific
fallibilism (Grenon and Smith, 2004). Thus, it is possible
that future research in ontology or in the natural sciences
will reveal the need for an expansion or restructuring of the
categories that BFO recognizes.

In its present form, BFO categories are those included in

the taxonomic hierarchy illustrated in Figures 1 and 2
above. However, we are exploring the possibility of intro-
ducing two sub-categories under function, namely artifac-

Arp & Smith

tual function and biological function, as illustrated in Figure
3.

We are also exploring the question of whether to include

tendency as a further sub-category within the ontology.

Fig. 3. Two proposed sub-categories of function in BFO.

BFO:realizable entity

function

artifactual function

biological function

role

disposition

3.1

Artifactual Function

An artifactual function is a function which inheres in an
independent continuant that exists, and has the physical
structure which it has, because it has been designed and
made intentionally (typically by one or more human beings)
to function in a certain way and does indeed reliably func-
tion in this way (Lind, 1994; Dipert, 1993).

Examples include: the function of a pycnometer to hold

liquid, the function of a fan to circulate air, and the function
of a Bunsen burner to produce a flame.

3.2

Biological Function

A biological function is a function which inheres in an inde-
pendent continuant that is (i) part of an organism and (ii)
exists and has the physical structure it has as a result of the
coordinated expression of that organism's structural genes
(Rosse and Mejino, 2003). The manifestations of a function
of this sort form part of the life of the organism.

Examples include: the function of a mitochondrion in the

production of ATP and the function of the wax-producing
mirror gland of the worker honey bee to produce beeswax.

The manifestations of biological functions are not in

every case beneficial to the survival of the corresponding
organism. (Consider the case of organisms that die when
they reproduce, like Arabis laevigata and Octopus lutens.)
Rather, they are (in typical environments) such as to contri-
bute to the realization by an organism of a life that is typical
or characteristic for an organism of its kind.

It is an open question whether the dichotomy between bi-

ological and artifactual function should or should not be
included as an addition to BFO, or reflected rather in the
creation of two new domain ontologies of artifactual and of
biological functions. The latter has already been proposed as
a complement to the GO's molecular function and biologi-
cal process ontologies.

ACKNOWLEDGEMENTS

We wish to thank Andrew Spear for helpful comments. This
work is funded by the United States National Institutes of
Health (NIH) through the NIH Roadmap for Medical Re-
search, Grant 1 U54 HG004028.

REFERENCES

Allen, C., Bekoff, M., and Lauder, G., eds. (1998) Nature's Pur-

poses: Analyses of Function and Design in Biology. MIT Press,
Cambridge.

Ariew, A. and Perlman, M. (2002) Introduction. In A. Ariew, R.

Cummins, and M. Perlman, eds., Functions: New Essays in the
Philosophy of Psychology and Biology (pp. 1

7).

Oxford Uni-

versity Press, New York.

Arp, R. (2006) Evolution and two popular proposals for the

definition of function. Journal for General Philosophy of
Science, 37, 212.

Dipert, R. (1993) Artifacts, Art Works, and Agency. Temple Uni-

versity Press, Philadelphia.

Grenon, P. and Smith, B. (2004) SNAP and SPAN: Towards dy-

namic spatial ontology. Spatial Cognition and Computation, 4,
99104.

Hill, D., Smith, B., McAndrews-Hill, M., and Blake, J. (2008)

Gene Ontology annotations: What they mean and where they
come from. BMC Bioinformatics, 9, S2.

Hunter, L. (2009) An Introduction to Molecular Biology for Com-

puter Scientists. MIT Press, Cambridge, MA. In preparation.

Jansen, L. (2007) Tendencies and other realizables in medical in-

formation sciences. Available at: http://ontology.buffalo.edu/
bfo/Tendencies.pdf.

Lind, M. (1994) Modeling goals and functions of complex indus-

trial plants. Applied Artificial Intelligence, 8, 259283.

Perlman, M. (2004) The modern philosophical resurrection of tele-

ology. The Monist, 87, 351.

Rosse, C. and Mejino, J. (2003) A reference ontology for bioin-

formatics: The Foundational Model of Anatomy. Journal of
Biomedical Informatics, 36, 478500.

Smith, B. and Grenon, P. (2004) The cornucopia of formal-

ontological relations. Dialectica, 58, 279296.

Smith, B., Kusnierczyk, W., Schober, D., and Ceusters, W. (2006)

Towards a reference terminology for ontology research and
development in the biomedical domain. Proceedings of KR-
MED, 1, 110.

Smith et al. (2007) The OBO Foundry: Coordinated evolution of

ontologies to support biomedical data integration. Nature Bio-
technology, 25, 12511255.