Final version published in S. Borgo, R. Ferrario, C. Masolo, and L. Vieu (eds.), Ontology
Makes Sense. Essays in Honor of Nicola Guarino, Amsterdam, IOS Press, 2019, pp. 3–23
Carnapian engineering
Achille C. Varzi
Department of Philosophy, Columbia University
Abstract. Ontology has come to gain huge currency in the information sciences,
with techniques, applications, and results vastly exceeding the traditional concerns
of philosophy. How did that happen? Where are we heading to? I do not have the
answers. But I know what it took and what is needed. It took—and we need—the
skills of a good Carnapian engineer, someone capable and willing to build bridges
across fields even though each side regards them as a troublesome intruder.
1. No ontology without Ontology
Ontology: the science of what is. While the word is of modern coinage,1 the field of
inquiry it denotes is as old as a science can be, indeed as old as philosophy. Yet ontology
is also, in many ways, a new discipline. For just as it developed over the centuries as
a central chapter of theoretical metaphysics, so in recent years ontology has become a
central concern in applied information science. Researchers in various fields have come
to realize that a solid foundation for their projects calls for an explicit theorization of
the types of entities and relations that make up their respective domains of inquiry, and
as the need to integrate such projects has grown, so has the need to identify common
general principles over ad hoc, case-based solutions—principles that have come to be
characterized as ‘ontological’.
It is difficult to locate the exact origins of this insight, but one of its earliest official
expressions—and one of the earliest formulations of the corresponding agenda—may be
traced back to the International Workshop on Formal Ontology in Conceptual Analysis
and Knowledge Representation held in Padua, Italy, in March 1993 on the initiative of
Nicola Guarino and Roberto Poli. It is worth quoting from their preface to the Proceedings volume:
Problems of conceptual analysis and of the acquisition of commonsense knowledge still represent a major bottleneck in the development of information systems. In mature areas like
databases as well as in more recent fields like knowledge engineering, the rigorous analysis
of the domain and the choice among the available representation primitives are difficult tasks,
1 The oldest extant records (in Latin and Greek, respectively) appear in two works of the early 17th century,
Jacob Lorhard’s Ogdoas Scholastica (1606) and Rudolf Göckel’s Lexicon philosophicum (1613) (see [18] for
details). The first occurrence of the term in English as recorded by the Oxford English Dictionary appears in
Harvey’s Archelogia philosophica nova (1663), which defines ontology as ‘a Discourse of a Being’, followed
by Bailey’s Dictionary (1721), which defines it as ‘an Account of being in the Abstract’.
1
appearing to rest on a mixture of introspection (as to what is intuitively right) and technical
realization of the resulting theory. Existing knowledge acquisition tools tend to be oriented
around applications, and therefore a new knowledge base must usually be constructed from
scratch when a new problem is undertaken. Due to the high costs of this process, recent initiatives like DARPA’s Knowledge Sharing Efforts have underlined the opportunity of increasing
the quality of formalized bodies of knowledge in such a way that it is possible to share and
reuse at least parts of them for a variety of different purposes. The development of so-called
generic ontologies plays an essential role in this connection. [41, p. i]
In fact, by the time of the Padua workshop, participants in the Knowledge Sharing initiative2 were already speaking of ‘ontologies’ in this sense:
If we could develop shared sets of explicitly defined terminology, sometimes called ontologies, we could begin to remove some of the arbitrary differences at the knowledge level.
Furthermore, shared ontologies could provide a basis for packaging knowledge modules—
describing the contents or services that are offered and their ontological commitments in a
composable, reusable form. [73, p. 38]
However, that was about it. Equating ontologies with sets of explicitly defined terminology was of no guidance, conceptually and methodologically, as to how those sets ought
to be put together; identifying the ontology of a system with ‘a vocabulary and a set
of constraints on the way terms can be combined to model a domain’ (p. 40) was of no
guidance as to how the relevant constraints should be determined. What emerged at the
Padua Workshop was precisely the need to go beyond such a crude characterization. The
development of composable, reusable ontologies cried for a methodology, a theoretical
framework, an ontological science in the good old sense. It is for this reason that while
the preface to the Proceedings emphasized the essential role of generic ‘ontologies’ (in
the plural), the title of the Workshop featured ‘ontology’ (in the singular). Moreover, the
emphasis in the title was on formal ontology, with a double reference to the meaning of
the adjective made popular by such contemporary authors as Nino Cocchiarella, on the
one hand, and the meaning inherited from Husserl’s classical conception, on the other.
The former is nearly tantamount to ‘rigorous’:
Formal ontology is the result of combining the intuitive, informal method of classical ontology
with the formal, mathematical method of modern symbolic logic, and ultimately identifying
them as different aspects of one and the same science. [14, p. 640].
Not much news there, particularly for applications in domains where formalization and
mathematical rigor are already common practice, if not a prerequisite. The latter understanding of ‘formal’, however, is much more significant, especially when combined with
‘ontology’. According to Husserl, formal ontology is not just ontology suitably formalized; it is ontology bereft of all domain-specific considerations, the science of what is, no
matter what it is, the study of the general ‘laws of being’, whose validity should possess
the same sort of universality and topic-neutrality that characterize the general ‘laws of
truths’ of formal logic:
The systematic place for its discussion [is] in the pure (a priori) theory of objects as such, in
which we deal with ideas pertinent to the category of object [. . . ] as well as the a priori truths
which relate to these. [54, p. 435]
2 Sponsored by DARPA but also, among others, by the National Science Foundation (SNF) and the Corporation for National Research Initiatives (CNRI).
2
Thus understood, formal ontology may not yet be enough of a ‘scientific’ framework
for developing specific ontological modules—what Husserl himself would call ‘material
ontologies’. One would need, in addition, a general theory of the sort of objects (entities)
that may feature in each module and of how they relate to one another—an ‘upper ontology’, in modern parlance. But that is already something that we can appreciate only
insofar as we engage with the picture from a broad perspective. And it was the need for
such a perspective that the Padua Workshop emphasized, setting an agenda that went far
beyond the scope of the Knowledge Sharing initiative and putting information scientists
in actual conversation with philosophers, de facto as well as de jure.
Since then, the project developed rapidly. Guarino himself played a pivotal role in establishing a major conference series to foster the conversation (FOIS – Formal Ontology
in Information Systems, launched in 19983 ) along with the first research institute devoted
entirely to the project (LOA – Laboratory for Applied Ontology, est. 20034 ), the first dedicated journal (Applied Ontology, est. 20055 ), and the first international consortium fully
aimed at promoting interdisciplinary research and collaboration (IAOA – International
Association for Ontology and its Applications, est. 20096 ). Many others have followed
in these footsteps, and today we can easily find handbooks [12, 72, 85, 92], textbooks
[20, 26, 56, 64], and collections [17, 58, 76, 93] providing comprehensive overviews
and detailed digests of this fast-evolving field. The range of applications is enormous,
spanning from conceptual modeling and knowledge representation to knowledge engineering, knowledge management, database design, object-oriented analysis, information
retrieval and extraction, query processing, natural language processing, machine translation, library science, enterprise integration, cloud computing, machine learning, agentbased systems design, general and precision medicine, healthcare simulation modeling,
biomedical informatics, genomic and nutritional epidemiology, chemical and pharmaceutical engineering, mechanical engineering, electronic commerce, geographic information systems, legal information systems, biological information systems, electronic
government, service science, the semantic web, the internet of things, and much more.
In 2012, the State of the Future report of the Millennium Project went as far as to include a new 51-page chapter (‘Future of Ontologists’) based entirely on the thought that
information-science ontology may soon have ‘an impact on humanity more profound
than the one caused by the Internet’ [25, p. 121]
No one of course may claim or take credit for all this. But it would not be an exaggeration to say that the picture would be quite different had the magic word, ‘ontology’,
been used merely as a fancy substitute for the crude notion of an ‘explicitly defined terminology’. Nor would the picture be so rich had the meaning of ‘ontology’ been limited
to its plural use. At the Padua Workshop, Tom Gruber presented a paper based on his
definition of an ontology as ‘an explicit specification of a conceptualization’ [29, p. 2].
The definition had just appeared in print elsewhere [28, p. 199] and would eventually
gain high currency among information scientists, in this form or slightly amended (e.g.
to include the requirement that the specified conceptualization be generally accepted, or
‘shared’, as urged by Willem Borst [11, p. 12]). No doubt this was a step up from the
crude notion elicited by the Knowledge Sharing initiative. But, again, it left pretty much
3 Website:
iaoa.org/fois
www.loa.istc.cnr.it
5 Website: www.iospress.nl/journal/applied-ontology
6 Website: iaoa.org
4 Website:
3
everything up for grabs. The real boost came from the realization that the very possibility
of sharing ontologies depends on developing them on the basis of shared criteria, and
that these criteria, in turn, can only come from a genuine engagement with the discipline
that comes with that name. In a slogan: no ontology without Ontology.
2. Welcome to the quagmire
With all this, no realization occurs alone. It didn’t take long to realize, too, that the project
would be everything but straightforward.
In the paper Guarino wrote for the special issue of the International Journal of
Human-Computer Studies following up the 1993 Workshop,7 the last section ended with
an invitation to ponder a passage from Drew McDermott’s review of Lenat and Guha’s
1990 book, Building Large Knowledge-Based Systems [61]:
Those were the good old days. l remember them well. Naive physics. Ontology for Liquids.
Commonsense Summer. It was a heady time, and this book rekindles the old excitement.
Wouldn’t it be neat if we could write down everything people know in a formal language?
Damn it, let’s give it a shot! [. . . ] We—and Lenat and Guha—are on a slippery slope here.
If we want to be able to represent anything, then we get further and further from the practicalities of frame organization, and deeper and deeper into the quagmire of logic and philosophy.
[67, pp. 54, 58]
Guarino’s concluding comment was candidly forthright: ‘I believe that this quagmire is
well worthwhile getting into’ [32, p. 637]. And so it has been. In the years that followed,
Guarino and his collaborators did not refrain from diving (and digging) into a number of
philosophical questions that the project immediately generated—questions concerning
e.g. the constitution of physical objects [9], the structure of space [8], the principles
governing identity, unity, and subsumption [35, 43, 45], the nature of properties [44],
the ontological status of roles and social objects [66], the distinction between events and
processes [37, 48], the part-whole relation [3, 4], and more. All of this while continuing
to promote the need and benefits of an ontology-based information science [31, 33, 34,
36, 39] and at the same developing a detailed upper-level ontology, DOLCE [24, 23, 65],
along with a rigorous methodology for analyzing and validating specific ontologies in
terms of formal meta-properties, OntoClean [46, 47, 99].
But the quagmire is deep and wide. The discipline of ontology is tortuous and full
of traps. Willard Quine famously emphasized one sort of trap when he focused on one
aspect of ontology, the aspect concerned with the material question ‘What is there?’ [79].
In a way, the answer is obvious: ‘Everything’. At least this is the obvious answer insofar
as existence is not expressed by an ordinary predicate but by the apparatus of quantification: to be is to be the value of a bound variable [78].8 There simply cannot be things
that do not exist. However, to say ‘Everything’ is to say nothing. As Quine hastened to
add, it is merely to say that there is what there is, an empty tautology, unless one goes on
to specify the contents of the domain over which one quantifies—and here there remains
plenty of room for disagreement. You may think that ‘everything’ covers particulars as
7 The special issue [42] contains 18 articles, 9 of which came from the 38 papers (including Gruber’s) comprising the Workshop proceedings.
8 Pace a certain philosophical tradition that insists on treating ‘exists’ as a predicate, inspired by Meinong’s
theory of objects [68] and variously defended in contemporary literature [6].
4
well as universals, I may think it only covers the former; you may think the domain of
quantification includes abstract particulars along with concrete ones, I may think it only
includes the latter; you may think the only concrete particulars are material bodies, I may
think they include immaterial entities as well, such as holes or shadows; and so on. Guarino’s digging into the quagmire led him to figure out his position on many such matters,
and everyone should do the same. But that’s hard business, and it’s bound to result in
different accounts.
And this is just the beginning. What exactly does it mean to disagree, ontologically
speaking? Suppose you and I disagree about whether there is, or is not, orange juice
in the refrigerator. Is that a case of ontological disagreement? Suppose we disagree on
whether there is a hole inside a chunk of cheese, or a rabbit-shaped shadow on the wall.
Ontological disagreement? Or suppose you and I agree on the existence of certain entities
but disagree on their nature—say, we agree that there are people but disagree on what
(as opposed to who) they are. You think people are enduring entities, things that persist
across time by being fully present at each time at which they actually exist; I think people
are perduring entities, which persist across time by virtue of having a distinct proper
part—a temporal part—for each time at which they exist. What are we to make of our
original agreement concerning people’s existence? Did we agree, ontologically speaking,
or did we not? And if we say we did, how and where is the relevant disagreement going
to show up in our theories, in our formalisms?
Similar worries apply to other questions that define the business of ontology. For
instance, we said that formal ontology—the science of what is, no matter what it is—
is supposed to consist of highly general, domain-independent principles, principles that
capture the fundamental ‘laws of being’. How do we figure out those laws? For Husserl,
they must be discovered a priori, but what does that mean, effectively? And what sort of
laws are they anyway? For Husserl, the pure theory of ‘objects as such’ must deal with
ideas ‘pertinent’ to the general category of object, ideas such as ‘Whole and Part, Subject
and Quality, Individual and Species, Genus and Species, Relation and Collection, Unity,
Number, Series, ordinal Number, Magnitude, etc.’ [54, p. 435]. Do we agree? How are
these ‘ideas’ to be selected, exactly? What exactly is it that makes them ‘formal’, as
opposed to others? How can we fill in the ‘etc.’?
Even before we can address any of these questions, there’s the big question of how to
interpret the ‘is’ of ontology. The science of what is; but what is ‘is’ supposed to mean?
Never mind looking for a substantive account of what it takes for something to be. (Does
being require spatio-temporal location? Does it require having causal powers? Etc.)9
Let’s stick to Quine’s quantificational account: to be is neither more nor less than to be
the value of a bound variable. Is the range of our variables meant to reflect our ontological
commitments, our beliefs, our conceptualization of the world, or is it meant to reflect
the content of the world itself? Is ontology about reality as we represent it, or is it about
reality as it is? This sort of question, and the opposition between conceptualism and
realism that it elicits, is responsible for some of the most profound and dramatic divides
in the history of philosophical ontology, and there’s no way out of it. We may wish to
ignore it for some purposes. But if we go for the quagmire, that is the first question that
is going to hit us.
9 And never mind Heidegger: “We understand the word ‘is’ (‘being’), we know the meaning; but we are
unable to say what we ‘really’ mean by it. We understand it, but we do not grasp it. We do not have a concept
of ‘is’. We understand ‘is’ and ‘being’, but in a non-conceptual way.” [51, p. 149]
5
None of these considerations is meant as big news. Anyone working in the field—
especially anyone who attended the FOIS conferences and kept abreast of the quarterly
issues of Applied Ontology—will find them familiar, perhaps even annoyingly so. Yet this
is not to say we can just move on. On the contrary, after twenty-five years of tremendous
growth and developments, this seems a good time to take stock and reopen the discussion,
at least to remind ourselves that it was never closed. Let me offer a taste of what this
would involve with reference to three aspects in which I think the questions raised above
are still worthwhile and, indeed, pressing.
3. Realism vs. conceptualism
Starting from the end, consider the opposition between a realist understanding of ontology and a non-realist, purely conceptualist one. We find statements of both views from
the very beginning. Gruber’s original account, for example, was clearly of the latter sort.
The definition recalled above continues thus:
An ontology is an explicit specification of a conceptualization. The term is borrowed from
philosophy, where an ontology is a systematic account of Existence. For knowledge-based
systems, what ‘exists’ is precisely that which can be represented. [28, p. 199; cf. 29, p. 2]
We are told what a ‘conceptualization’ is, namely, an ‘abstract, simplified view of the
world that we wish to represent for some purpose’. But there is no claim whatsoever as
to how a view and the corresponding knowledge base relate to the underlying world.
In particular, there is no reference to the idea that knowledge, unlike belief, should be
factive, i.e., truth-entailing.10 By contrast, Guarino’s approach was driven precisely by
that idea, with all the realism that comes with it:
A knowledge base will acquire a value per se only to the extent that the knowledge it contains
is in fact true, such as to correspond to the world beyond the knowledge base. Therefore,
the study of ontology, intended as a branch of philosophy dealing with the a priori nature
of reality, can be of benefit to the knowledge-construction process in yielding high-value
knowledge bases. [32, p. 626]
Of course there is a difference in the way the word ‘ontology’ was used in the two cases.
Gruber was using it as a count noun, telling us what an ontology is; Guarino was using
‘ontology’ as a proper name, the name of the philosophical discipline (elsewhere written
with uppercase initial). We already emphasized that the difference is important. And it
surely is in this connection. Indeed, when it comes to the countable use, Guarino himself would not refrain from defining ontologies in terms of ‘conceptualizations’ [38, 40].
Nonetheless the emphasis on the underlying reality and its a priori nature was clearly different, reflecting a different approach in the way we should go about developing reusable
ontological modules. Gruber would eventually settle on purely pragmatic guidelines:
One can say that ontology is a tool and product of engineering and thereby defined by its use.
From this perspective, what matters is the use of ontologies to provide the representational
machinery with which to instantiate domain models in knowledge bases, make queries to
knowledge-based services, and represent the results of calling such services. [30, p. 1964]
10 This is—and was—in line with standard practice in the AI field of so-called knowledge representation. It’s
unfortunate and by now it’s a lost battle, but ‘belief representation’ would have been a more appropriate label;
see e.g. [82].
6
Guarino would insist that use and reusability go hand in hand and should be driven by a
systematic effort to understand every conceptualization, even when resulting ‘from our
sensory system, our culture, and so on’ [45, p. 113], in terms of the general structure of
the underlying reality—the thesis of ‘the independence of domain knowledge’:
Reusability across multiple tasks or methods can and should be systematically pursued even
when modeling knowledge related to a single task or method: the more this reusability is
pursued, the closer we get to the intrinsic, task-independent aspects of a given piece of reality
(at least, in the commonsense perception of a human agent). In this systematic quest for
reusability, the potential role of a discipline like formal ontology appears evident. [33, p. 294]
The tension between these different approaches has of course different strengths and
implications precisely inasmuch as we think it’s worth getting into the quagmire. For a
while it was taken seriously, setting the agenda of an intense debate. Barry Smith, for
one, emphatically pressed the need to get clear about the role of reality and reference
thereof, fearing that the knowledge-representation and related communities, when doing
ontology, might ultimately embrace ‘one or other form of idealist, skeptical, or constructionist philosophy’ [89, p. 73]. This led to a vehement exchange on the pages of Applied
Ontology, with Smith and collaborator Werner Ceusters defending the realist approach
[90] and Garry Merrill questioning it [69, 70]. The exchange was felt to be so climacteric
that IOS Press, the journal’s publisher, decided to make the articles available online free
of charge and without restrictions, issuing two press releases to invite ‘everyone who
cares about principles of ontology engineering, about what constitutes a good ontology,
and about the future of ontology in e-science’ to ponder this ‘much important, unsettled business’.11 A remarkable event indeed, promptly followed by hundreds of postings
on many online fora. But, alas, the quagmire really is deep and wide. And after all the
excitement, the topic was gently dropped and the business remains unsettled.
Why is it important to settle it? Not only because of its intrinsic interest. One important reason, which no information-science ontologist can swiftly brush aside, is that
the alternative between realism and conceptualism tends to translate into a parallel alternative between monism and pluralism in scientific practice, i.e., between the view that
there is just one way of getting things right versus the view that many ways—even incommensurable or mutually incompatible ways—can be equally correct. No doubt we
can have a plurality of ontologies, which is to say, many ways of building an ontology for
specific purposes, and it is a fact that such ontologies may not agree even on matters of
common concern. Otherwise the quest for sharing and reusability would be a cinch. But
what about the higher-level guidelines? Should each ontology conform to the same set of
general, formal principles? More concretely: should each domain ontology conform to
the same upper-level ontology, providing that the latter is precisely the sort of theoretical
framework that can benefit from the discipline of Ontology?
The realist has a swift and straight answer: Yes. It is true that, as John Searle put it,
‘realism does not say how things are but only that there’s a way that they are’ [84, p. 155].
Yet this is enough to warrant a strong form of scientific monism, generally as well as
in the present context. For all the legitimacy of the many domain ontologies that may
flourish, reflecting our first-order ontological disagreements, a realist attitude towards
Ontology will nonetheless aim at delivering ‘a definitive and exhaustive classification of
entities in all spheres of being’ [88, p. 155]. We may disagree on what there is, or on
11 Both
press releases are still available at http://www.applied-ontology.org/ontologicalrealism.
7
various other matters of contingent fact. But the laws of being are not up for grabs. They
are what they are, determining a kind of ‘generalized chemistry’ [91, p. iii] to which
everything must conform. By contrast, a conceptualist is hardly going to push this line.
I say ‘hardly’ because strictly speaking conceptualism does not entail world pluralism.
A lot depends on the details of one’s semantic theory, especially with regard to reference
and truth.12 But details such as these can only be found deep down in the quagmire. As
a matter of general attitude, conceptualists don’t care about the world; they care about
conceptualizations. And the laws of conceptualizations are up for grabs.
So why is this important? I mentioned that among the many achievements of Guarino and his research group is the development of a rich and powerful upper ontology,
DOLCE . The principles of DOLCE have been developed explicitly ‘with the vision that a
unique universal ontology for knowledge representation cannot exist’ [10, p. 372]. But
DOLCE is not the only product of this sort. Several other upper ontologies have been developed over the years, from CYC [60] and BORO [75] to SUMO [74], BFO [2], UFO [49],
GFO [52], and many others (including YAMATO [71], short for ‘Yet Another More Advanced Top-level Ontology’). Such abundance of systems is by itself indicative of how
much the field has grown since the early days. But each system deviates from the others
in significant ways, not only by employing different methods and techniques but also,
crucially, by relying on different ontological categories and principles. And herein lies
the trouble. If the main reason to embark on ontology-based information science was to
overcome the ‘database Tower of Babel’ problem, as Smith called it [88, p. 158], after
all this work we seem to find ourselves with a parallel problem at a higher level—what
Wacław Kuśnierczyk calls the ‘ontology Tower of Babel’ problem [59, p. 41]. Quis custodiet ipsos custodes? The monist realist will insist on their bold answer: the supreme
custodian is the World, like it or not. But the conceptualist? This is not just an abstract
philosophical question. It is as concrete and pressing, from an information-science perspective, as the start-off question raised by Knowledge Sharing initiative.
4. The whence and whither of disagreement
The second issue concerns the very notion of ontological agreement. I mentioned the
possibility that you and I agree that people exist but disagree on what they are: you
think people are enduring entities, I think they are perduring entities. Obviously the case
generalizes. With regard to ordinary objects, too, you and I may agree on their existence
while holding dffferent views concerning their nature and identity conditions. I may think
a house is merely an aggregate of walls and roof; you may think it is something else,
something more. You may think a statue is constituted by its matter; I may think the statue
is the matter. And what goes for people and ordinary objects goes for everything—events,
properties, relations, functions, roles, groups, musical compositions, fictional characters,
mathematical entities, angels, what have you. In each of these cases, it would seem that
we may find ourselves agreeing on whether something is despite disagreeing on what it
is. Hence the question: would these be cases where you and I can be said to share an
ontology, albeit perhaps imperfectly?
12 Indeed, one could make a similar point about realism: strictly speaking it does not entail monism. For
example, in the philosophy of biology there is a well-respected view (known as ‘pluralistic realism’) that tries
to reconcile realism about species with taxonomic pluralism [19, 57].
8
Philosophically, this is a deeply controversial question [97]. Yet, for all their differences, virtually every theory developed from an information-science perspective delivers,
or tends to deliver, the same answer: No.
This is certainly true of those theories that follow Gruber’s definition of ‘ontology’
(or variants thereof). If an ontology is a specification of a ‘conceptualization’, and if a
conceptualization, in turn, is a ‘view of the world’, then it may seem as though we agree
on what there is but, on closer look, our ontologies diverge. On closer look, we may agree
that, say, the sentence ‘People exist’ is true. But if we hold a different conception of what
people are, then you and I are attaching different meanings to the word ‘people’ and,
therefore, to the whole sentence. You take the sentence to express the proposition that
people1 (enduring-people) exist; I take it to express the proposition that people2 (perduring-people) exist. Since people1 and people2 are different sorts of things, the commitments we incur in asserting the sentence are different, and hence so are our ontologies.
However, the point is not limited to Gruberian ontologies. Every ontology, insofar as
it is governed by suitable upper-level principles, will yield the same verdict. The reason
is that every upper-level ontology is bound to include a subsumption relation of sorts,
an ‘is a’ relation, and this will suffice to associate every item in the domain with a characterization of sorts. The ‘is a’ taxonomy will classify every thing according to what
it is. At least, it will do so in terms of the categories available in the language, which
is precisely why an ontology informed by the philosophical discipline that goes by that
name is supposed to be better, and ultimately more informative, than one informed by
loose intuitions and common-sense distinctions. Upper ontologies are not just inventories of the contents of our domain of quantification; they are structured inventories. And
if we disagree on the structure, we disagree on the whole. (In this sense, the relevant
notion of ‘ontology’ differs from Quine’s as well as from Husserl’s, reflecting instead the
conception of Roman Ingarden and his school [55].)
Now, by itself there is nothing wrong with this picture. But it’s worth pondering its
presumptions as well as its consequences. To begin with the former, do we really want
to rule out the possibility of mere ontological disagreement? Surely, generally speaking
there is nothing incoherent in the idea that two parties may have different opinions concerning one and the same entity. This is obvious in the case of ordinary opinions. For you
the statue of David is splendid; for me it isn’t. It doesn’t follow that we are referring to
two distinct objects—one splendid and one unimpressive. For you Cattelan is an artist;
for me he is not. It doesn’t follow that we are referring to two Cattelans. The same may
be said of contrasting opinions in scientific contexts. Renaissance people used to think
that water was one of the simple elements of the sublunar world; with Lavoisier we came
to recognize that water is instead a compound of hydrogen and oxygen. It doesn’t follow
that the Renaissancers thought they were drinking something else than what we think we
are drinking. Otherwise the very notion of scientific progress would become empty, if
not meaningless.13 Well, then, the same might be said of contrasting opinions of the sort
illustrated above—call them metaphysical opinions. You and I have different conceptions of people? It doesn’t follow that we are speaking of different things; we may sincerely disagree on what people are. And if an oracle told us that the correct metaphysic
is endurantist, contrary to what I thought, I would not react by denying the existence of
people. Simply, I would accept the news and revise my conception accordingly. Isn’t this
13 This is one important lesson of Hilary Putnam’s treatment of scientific terms (and theories) in ‘The Meaning of ‘Meaning” [77].
9
what we all try to achieve when we argue with our colleagues? We all try to make them
change their views on a certain subject, not to change the subject.
All of this is ruled out from start in the way domain ontologies are typically governed
by upper-level ontologies. And there is more. For not only is there no incoherence in
the idea that two parties may hold different metaphysical opinions concerning the same
entities; there is nothing incoherent even in the idea that we may proclaim our ontological
credo without embarking at all on metaphysical speculations. Consider what happens in
mathematics. Most mathematicians would avow that they are serious about numbers—
that numbers are indeed among the values of their bound variables. Yet few would be
willing to go further and say something committal about the nature of such values—about
what numbers are. How are we going to model the beliefs of such agents? Are we going
to fill in some details on their behalf, insisting e.g. on the abstract nature of numbers? Are
we going to identify numbers with Fregean extensions of second-level concepts? With
cumulative sets? Which ones—the Zermelo sequence, ∅, {∅}, {{∅}}, {{{∅}}}, . . . ,
or the von Neumann sequence, ∅, {∅}, {∅, {∅}}, {∅, {∅}, {∅, {∅}}}, . . . ? Surely we
have several options, and if we look at philosophical Ontology we may find more. But
it seems presumptuous, indeed ludicrous, to represent our mathematicians as holding
such views. And mathematics is but one example. Many biologists, for instance, feel
the same way in regard to the metaphysical disputes about species (e.g. whether species
are classes or individuals). The same may apply to ordinary beliefs concerning people,
houses, statues, etc. I can hear my neighbor: ‘People? Of course people exist! But don’t
ask me whether they endure or perdure, I don’t even know what that means and I don’t
care. I believe in people whatever they are.’
One might protest that merely asserting the existence of something without saying
what it is would be empty talk. To say that people exist ‘whatever they are’ would be
completely uninformative, and similarly for species, numbers, etc. However, that is not
what happens in such cases. In the case of numbers, mathematicians do not leave us entirely in the dark: whatever they are, numbers must enjoy a specific range of attributes,
say, those attributes that define the axiomatic corpus of Peano arithmetic. In saying that
numbers exist, therefore, mathematicians are committing themselves to the existence of
entities that enjoy those attributes; their commitment is perfectly clear.14 Likewise, in the
case of entities such as people, houses, or statues. There is a whole range of attributes
that such things are supposed to have in virtue of the application conditions, or ‘meaning
postulates’, that come with the correct use of the relevant words—attributes such as being
featherless, biped, and somewhat rational (people), having a certain structure and serving
the purpose of providing shelter (house), being the product of sculpturing, modeling, or
casting (statue), and so on. Such postulates reflect a cluster of common-sense truths and
do not, therefore, constitute a rigorous axiomatic theory comparable to Peano arithmetic
(though early AI projects in naive physics [50] and the commonsense world [53]—the
‘good old days’ evoked by McDermott—had precisely that ambition). Still, such postulates are constitutive of our linguistic competence and suffice to fix the extensions of our
predicates. And it is precisely because my neighbor is linguistically competent that we
understand her when she says ‘People exist’.
So much for the presumption worry: our upper-level ontologies appear to build too
much metaphysics into our domain ontologies. Indeed, here the project of a philosophy14 That is, clear ‘up to isomorphism’, as some like to say. I take this to be the main lesson of Paul Benacerraf’s
‘What Numbers Could not Be’ [5].
10
inspired approach risks backfiring. The more they are inspired by the philosophical discipline of Ontology, rather than by ordinary dictionaries, the higher the risk of going
overboard. But there is something to be said also about the practical consequences of this
state of affairs. Specifically, it’s clear that ruling out from the start the possibility of mere
ontological disagreement multiplies the challenges of integration and interoperability,
resulting in greater chaos in the ontology Tower of Babel.
So this is our second example of a philosophical issue worth pondering. One suggestion would be to rethink the way we tend to structure our top-level distinctions. We
have seen that there are two criteria for taxonomizing things (or specifying concepts): according to what they are (or mean) insofar as this is required for reference fixing, and according to some deeper, full-bodied metaphysical categories. In both cases, the relevant
structural relation is a form of subsumption. But it need not be the same. We could split
the ‘is a’ relation into two, one governed by meaning postulates, the other governed by
metaphysical axioms. The former would constitute the lexical, or terminological, module
of the theory; the latter its metaphysical module. And we can try to work out effective
ways of maximizing sharing on the basis of the corresponding forms of agreement that
may apply across ontologies: mere ontological agreement and full-bodied agreement, respectively. But it’s easier said than done. And, of course, drawing a clear line between the
two sorts of subsumption relations would not be easy—possibly a source of yet another
Babel Tower.
5. The formal sphere
The third and final point concerns the idea that in building an ontology we can benefit especially from the discipline of formal ontology. As noted, this idea has been the hallmark
of Guarino’s project from the beginning, to the point of using the adjective, ‘formal’,
both in the title of the 1993 Workshop (Formal Ontology in Conceptual Analysis and
Knowledge Representation) and in the official denomination of the regular conference
series established five years later (Formal Ontology in Information Systems). And while
the sense of ‘formal’ that is synonymous with ‘rigorous’ is obviously important, it is
especially in the sense inherited from Husserl that the idea proved powerful in relation
to the quest for information sharing and reusability. As a science of what is, no matter
what it is, Husserlian ontology provides us with tools that should be applicable across the
board, no matter the specific motivations and tasks of each particular knowledge base.
Regardless of whether the domain includes objects along with events, concrete entities
along with abstract ones, and so on, it must exhibit some general features and obey some
general laws,. And ontology—understood formally—is the science of such features and
laws, a sort of ‘logic of being’.
Indeed, the use of ‘logic’ here is more than metaphorical. For Husserl, formal ontology deals with the ‘interconnections of things’ exactly as formal logic deals with the
‘interconnections of truths’ [54, p. 225]. Just as formal logic deals with laws that are formal in the sense that they apply to any proposition, no matter what it says, so the laws
of formal ontology are formal in the sense that they apply to any entity, no matter what
it is. Just as, for instance, the transitivity of entailment (if p entails q and q entails r, then
p entails r) does not depend on what propositions are assigned to the variables ‘p’, ‘q’,
and ‘r’, so the transitivity of, say, identity (if x is identical to y and y is identical to z,
11
then x is identical to z) does not depend on what entities are assigned to the variables ‘x’,
‘y’, and ‘z’. Both laws exhibit the same sort of generality and topic-neutrality. Both are
meant to hold as a matter of necessity and should be discovered, in some sense, a priori.
One can see why Guarino was invigorated by the project. And one can see why the
first edition of FOIS opened with a plenary talk by Barry Smith devoted precisely to
clarifying the formal logic-ontology parallel [87]. In the field of Knowledge Representation, formal logic was the tool (so much so that the full denomination of the field is actually ‘Knowledge Representation and Reasoning’). Adding formal ontology as a second
tool would require some homework, but it could literally double the potential. The point
I want to stress, with reference to the brief remarks made earlier, is that despite the elegance of Husserl’s parallel, the exact scope of what counts as formal ontology remains to
be determined. So it’s not just a matter of doing the homework; one needs to do serious
work, work that even philosophers have come to recognize only recently.15
The issue is really two-sided. One side concerns the bounds of the theory, the
other its contents. With regard to the first side, earlier I quoted Husserl’s list of formalontological relations, which includes Whole and Part, i,,e., dependence and parthood,
and I have just mentioned identity as another obvious example. But of course one need be
more precise. Are these truly formal in the intended sense, i.e., do they all apply to anything that might conceivably exist, no matter what it is? And are they the only relations
of this kind?
The formal character of the identity relation can hardly be questioned. As Quine
famously put it, no entity without identity [80, p. 20]. In fact, it is precisely because it
is absolutely general and domain-independent that identity is often treated as a formal
logical relation. Quine himself was firm about this, emphasizing that identity ‘knows no
preference’; it ‘treats of all objects impartially’ [81, p. 62]. Nevertheless, precisely because it is an objectual relation rather than a sentential operator—because it relates things
in the world rather than truths about the world—it is best to treat identity as pertaining
to formal ontology, not logic.
The formal character of dependence is perhaps equally unquestionable. Generally
speaking, the idea is that x depends on y if, and only if, x could not exist without y.16
There may be worries concerning the exact meaning of the modal locution ‘could not
exist’. However, because existence is not a predicate, there is no restriction on the possible values of x and y, which range over the entire domain of quantification. So, again, it
would appear that the relation of dependence enjoys the right sort of generality.
But what about parthood? This is perhaps the most common relation, along with
subsumption, to feature in the upper-level ontologies that took Guarino’s advice seriously, including Guarino’s own DOLCE, and indeed most philosophers would agree that
this relation applies not only to material objects, or to entities located in space and time,
but to all entities whatever—that it is topic-neutral and thus applies across all ontological categories. Others, however, disagree. For example, the thought that there are mereologically structured universals is sometimes found to be problematic. In David Lewis’s
example [62, p. 34], each methane molecule consists of one carbon atom and four hydrogen atoms. Are the universals carbon and hydrogen part of the universal methane? If they
aren’t, does that mean that the three universals (as opposed to their instances) have nothing in common? If they are, does it follow that hydrogen is part of methane four times
15 The
16 This
remarks that follow draw on the second part of [96].
is actually a simplification and is open to counterexamples [21], but we can skip the details.
12
over? What could that possibly mean? How could one thing be part of another more than
once? Of course, if we agree that parthood includes identity as a limit case (so-called
‘improper parthood’), we may always treat all problematic cases as mereological atoms,
things that have no further parts except for themselves. That would be enough to warrant the claim that parthood knows no restriction. But it’s a slippery move. By the same
pattern, one might extend any relation R to R-plus-identity and treat all R-problematic
cases as atomic in the relevant sense. That would hardly be a reason to treat the extended
reflexive relation as a piece of formal ontology in addition to identity.
Similar remarks apply to the other relations included by Husserl’s list, though this
example should suffice to highlight the difficulty of the task. On the other hand, surely
there may be other, non-artificial relations that do fit the bill. To fix the scope of formal
ontology we must rule out the bad candidates but we must also make sure to rule in
all the good candidates. Popular examples would include membership, inherence, and
connectedness, all of which have been studied extensively. I do not intend to settle the
question here. Just as logicians have been having a hard time figuring out a good way
of demarcating the bounds of logic, demarcating the bounds of ontology, in the formal
sense of the term, is no straightforward business. What I want to stress, rather, is that
here the difficulty may depend at least in part on our ontological beliefs, now in the
Quinean, material sense of ‘ontological’. We are looking for relations that are topicneutral and take absolutely all possible entities as arguments, and that requires unlimited
open-mindedness. Who’s got the gift? Surely there may be more things in heaven and
earth than are dreamt of in our philosophies.17 If nobody had ever dreamt of such things
as universals, let alone recognized their ontological dignity, the above worry concerning
the formal character of parthood would not have arisen. If all we had dreamt of were
material substances, a cornucopia of spatiotemporal relations would have qualified as
formal in the relevant sense. And so on. There is, in fact, a hidden quantifier in the
characterization of what counts as formal, a quantifier ranging over all possible (i.e.,
conceivable) entities. And it is by no means clear that we can grasp its range without
engaging in material-ontological considerations.
This, then, is one side of the issue: what are the exact bounds of formal ontology?
The other side concerns the contents of the theory. Regardless of how far it extends,
the truths of formal ontology are supposed to possess the same sort of generality and
topic-neutrality that characterizes the truths of formal logic. Yet, as soon as we start
digging, we realize that this characterization is very hard to pin down without begging
the question.
Consider identity. Surely not every identity-theoretic principle qualifies as formalontological in the intended sense. We must draw a line somewhere. For instance, few
are willing to endorse the principle known as identity of indiscernibles: it fails in some
possible worlds (e.g. a perfectly symmetric world [7]), if not in this world of ours (as
quantum mechanics would seem to suggest [22]). The converse principle, the indiscernibility of identicals, is more robust and certainly less controversial, but even that has
been found problematic in some contexts, e.g., vis-à-vis the phenomenon of qualitative
change. Drawing a line is always difficult. And in this case it seems clear that the difficulty depends once again on a careful consideration of what there is, here or in some
other possible world. Perhaps we should just stick to the very basics: identity is reflexive,
17 As there are philosophies that have dreamt of things that are neither on earth nor in heaven. This is a
separate issue, but not entirely off-topic [95].
13
transitive, and symmetric—an equivalence relation. Yet these properties have been questioned, too. For instance, you don’t have to be a hard-core dialetheist to think that there
are or might be non-self-identical objects; it might suffice to consider again the elusive
citizens of the quantum world, as Schrödinger famously argued [83]. We may find that
utterly absurd. But it is hard to give expression to our feelings while claiming full and
unbiased neutrality.
Or consider parthood. Few would be willing to buy into the whole body of classical
extensional mereology [16]. Some of its basic principles, such as extensionality and unrestricted composition, are highly controversial, and treating them as expressing formalontological truths would be missing the point. Such principles are best construed as expressing specific views on how things are, even on what things there are. Nelson Goodman [27], for instance, took mereological extensionality to be the hallmark of a nominalist stance—hence of a very precise material-ontological view. Likewise, to accept
unrestricted composition is to countenance the existence of a fusion for any non-empty
collection of things—something that has all those things as parts and has no part that is
disjoint from each of them. Perhaps the fusion is nothing over and above the things that
compose it, as David Lewis emphatically argued [63]. But to the extent that it qualifies
as something else than those things, it is clear that our attitude towards this principle is
bound to reflect our domain ontology.
So what mereological principles do we have in mind, insofar as parthood is supposed to be a formal-ontological relation? A standard candidate is the principle known
as weak supplementation, to the effect that no entity can be comprised of a single proper
part. Peter Simons [86] regards this as a bare minimum that we can require of a relation, along with reflexivity, antisymmetry, and transitivity, if it is to count as parthood at
all.18 Yet even here there is room for disagreement. Possible counterexamples include,
for instance, recent theories of material constitution, which hold that a material object
(a statue) and the matter that constitutes it (a lump of clay) are proper parts of each other
although neither has parts disjoint from the other [94]. One may be inclined to dismiss
such counterexamples as unintelligible precisely because they violate weak supplementation, but one might as well go the other way around and regard the independent plausibility of those theories as evidence against the principle. How can we settle the issue
without begging the question, if not by resorting to material-ontological considerations
of some sort?
Even the partial-ordering axioms have sometimes been disputed. The antisymmetry
of parthood, for instance, is immediately challenged by the constitution theories just
mentioned (the statue and the clay are part of each other but non-identical). But it could
be argued that the axiom is too strong regardless: in view of certain developments in
non-well-founded set theory (i.e., set theory tolerating cases of self-membership and,
more generally, of membership circularities [1]), one might for instance suggest building
mereology on the basis of an equally less restrictive notion of parthood that allows for
closed loops [15]. Such a suggestion could hardly be dismissed if we are insisting on
the formal status of parthood: if sets have to have a mereological structure, it is natural
to identify the parts of a set with its subsets. So, either we refuse to countenance nonwell-founded sets—and that is a straightforward claim about what there is—or we must
concede that such sets violate antisymmetry.
18 At least insofar as parthood is construed broadly to include identity as a limit case; proper parthood would
be transitive but irreflexive and asymmetric.
14
These are just examples. Nonetheless, for parthood as well as for identity (and the
case for dependence is not different) they ought to be indicative of how difficult it is to
come up with good, neutral criteria for drawing a line between purely formal principles
and substantive theses. Of course we have a similar problem in logic. Sometimes a logical principle is challenged on the grounds of a disagreement concerning the meaning of
certain logical operators. This is the case, for instance, with the principle of double negation in intuitionistic logic, or disjunctive syllogism in some relevant logics. In such cases,
perhaps Quine’s attitude says it all: change of logic, change of subject [81, p. 80]. In other
cases, however, the disagreement has nothing to do with matters of meaning; it concerns
precisely the material-ontological neutrality of the principles in question [98]. Think of
the controversies on the existential presuppositions of subalternation in Aristotle’s syllogistic, or of universal instantiation in contemporary predicate logic. Think of the failure
of distributivity in quantum logic. Think of the problematic status of the Barcan formulas in modal predicate logic. Even the most fundamental principles of classical logic,
such as the law of non-contradiction or the law of excluded middle, have sometimes been
questioned on such grounds: that there are no inconsistent facts, or that every fact is fully
determinate, appear to be claims that reflect explicit material-ontological commitments.
It should not be surprising, therefore, that the same sort of worry arises when we shift
our attention from the general theory of truths as such to the general theory of objects as
such, which is to say from formal logic to formal ontology. But that is no excuse. If we
want to use formal ontology, we need to figure out what it is.
6. The Carnapian engineer
All of this is meant to illustrate the need (and practical importance) to continue pondering the deep challenges that come with the project of developing good, well-founded
ontology-based information systems. It’s not just that it is worth getting into the quagmire, as Guarino nicely put it in the passage quoted earlier; it is necessary that we do
so, or else we might as well learn to live in the database Tower of Babel, with its own
morass of idiosyncrasies and ad hoc solutions. Precisely because my goal is illustrative,
I have no conclusions to offer. But I’d like to add one final comment, concerning the
spirit rather than the letter of these challenges. And it is entirely ad personam.
In October 2014, Nicola Guarino delivered a lecture at Carnegie Mellon University
entitled ‘Twenty Years of Applied Ontology’. As the title suggests, the lecture was intended as a state-of-the-art overview of applied ontology, taking the Padua Workshop
as the official birth date of the discipline. The abstract, still available online,19 is worth
quoting in its entirety.
More than 20 years have passed since the first international workshop on formal ontology
and information systems held in Padova, Italy, in March 1993. What is now called Applied
Ontology is an established interdisciplinary area of research which builds on philosophy, cognitive science, linguistics and logic with the purpose of understanding, clarifying, making explicit and communicating people’s assumptions about the nature and structure of the world.
This orientation towards helping people understanding each other distinguishes applied ontology from philosophical ontology, and motivates its unavoidable interdisciplinary nature.
In this talk I will review—under a personal perspective—the main achievements of Applied
19 At
www.cs.cmu.edu/calendar/fri-2014-10-31-1000/cmu-ontology-studies-and-engineering-seminar-cose.
15
Ontology in the last 20 years and the open challenges we are facing today, focusing in particular on the methodology of ontological analysis and on its applications in the context of
information systems design.
This is short and swift, and it is exactly what you would expect from the abstract of such
a lecture. But there is also a hint to a ‘personal perspective’, suggesting the lecture might
contain more than ordinary business. I was not present, so I cannot say. However I found
the slides of the talk on the internet.20 I looked for the personal perspective, and at first
sight it seemed as though it was left entirely to the verbal part of the lecture; the slides are
full of diagrams and data. But then I looked again, and I noticed that somewhere towards
the middle (slide 43), there is a quotation from Carnap’s ‘Intellectual Autobiography’.
That, too, is worth quoting in full:
If one is interested in the relations between fields which, according to customary academic divisions, belong to different departments, then he will not be welcomed as a builder of bridges,
as he might have expected, but will rather be regarded by both sides as an outsider and troublesome intruder. [13, p. 11]
This may well be the most elegant way I have ever seen to add a personal touch to a
professional lecture without sounding too personal or upfront. And it says it all. It says
the passion, the difficulties, the real challenges that come with this sort of work. If there
are any Carnapian engineers out there, and if they ever decide to form a society, Nicola
will, I hope, receive the honorific he deserves.21
References
[1] P. Aczel, Non-Well-Founded Sets, Stanford (CA), CSLI Publications, 1988.
[2] R. Arp, B. Smith, and A. D. Spear, Building Ontologies with Basic Formal Ontology, Cambridge (MA), MIT Press, 2015.
[3] A. Artale, E. Franconi, N. Guarino, and L. Pazzi, ‘Part-Whole Relations in ObjectCentered Systems: An Overview’, Data & Knowledge Engineering, 20 (1996),
347–383.
[4] A. Artale, N. Guarino, and M. C. Keet, ‘Formalising Temporal Constraints on
Part-Whole Relations’, in G. Brewka and J. Lang (eds.), Principles of Knowledge
Representation and Reasoning. Proceedings of the 11th International Conference
(KR’08), Menlo Park (CA), AAAI Press, 2008, pp. 673–683.
[5] P. Benacerraf, ‘What Numbers Could not Be’, Philosophical Review, 74 (1965),
47–73.
[6] F. Berto, Existence as a Real Property. The Ontology of Meinongianism, Dordrecht,
Springer, 2013.
[7] M. Black, ‘The Identity of Indiscernibles’, Mind, 61 (1952), 152–164.
[8] S. Borgo, N. Guarino, and C. Masolo, ‘A Pointless Theory of Space based on Strong
Connection and Congruence’, in L. Carlucci Aiello, J. Doyle, and S. C. Shapiro
(eds.), Principles of Knowledge Representation and Reasoning. Proceedings of the
5th International Conference (KR’96), Boston (MA), Morgan Kaufmann, 1996,
pp. 220–229.
20 Also available online, www.slideshare.net/NicolaGuarino1/21-years-of-applied-ontology, though here the
title says—more accurately—‘Twentyone years’.
21 Many thanks to the referees for their helpful comments on an earlier draft of this paper.
16
[9] S. Borgo, N. Guarino, and C. Masolo, ‘An Ontological Theory of Physical Objects’, in L. Ironi (ed.), Qualitative Reasoning. Proceedings of the 11th International Workshop, Pavia, IAN-CNR, 1997, pp. 223–231.
[10] S. Borgo and C. Masolo, ‘Foundational Choices in DOLCE’, in the second edition
of [92], pp. 361–381.
[11] W. N. Borst, Construction of Engineering Ontologies for Knowledge Sharing and
Reuse (Ph.D. dissertation), Enschede, University of Twente, 1997.
[12] C. Calero, F. Ruiz, and M. Piattini (eds.), Ontologies for Software Engineering and
Software Technology, Berlin, Springer, 2006.
[13] R. Carnap, ‘Intellectual Autobiography’, in P. A. Schilpp (ed.), The Philosophy of
Rudolf Carnap, La Salle (IL), Open Court, 1963, pp. 1–84.
[14] N. B. Cocchiarella, 1991, ‘Ontology II: Formal Ontology’, in H. Burkhardt and B.
Smith (eds.), Handbook of Metaphysics and Ontology, Munich, Philosophia, 1991,
pp. 640–647.
[15] A. J. Cotnoir and A. Bacon, ‘Non-Wellfounded Mereology’, Review of Symbolic
Logic, 5 (2012), 187–204.
[16] A. J. Cotnoir and A. C. Varzi, Mereology, Oxford, Oxford University Press, forthcoming.
[17] J. Davies, R. Studer, and P. Warren (eds.), Semantic Web Technologies: Trends and
Research in Ontology-based Systems, Chichester, Wiley, 2006.
[18] M. Devaux and M. Lamanna, ‘The Rise and Early History of the Term Ontology
(1606-1730)’, Quaestio, 9 (2009), 173–208.
[19] J. Dupré, ‘Natural Kinds and Biological Taxa’, Philosophical Review, 90 (1981),
66–90.
[20] D. Fensel, Ontologies. A Silver Bullet for Knowledge Management and Electronic
Commerce, Berlin, Springer, 2001.
[21] K. Fine, ‘Ontological Dependence’, Proceedings of the Aristotelian Society, 95
(1995), 269–290.
[22] S. French, ‘Why the Principle of the Identity of Indiscernibles is not Contingently
True Either’, Synthese, 78 (1989), 141–166.
[23] A. Gangemi, N. Guarino, C. Masolo, and A. Oltramari, ‘Sweetening WordNet with
DOLCE’, AI Magazine, 24/3 (2003), 13–24.
[24] A. Gangemi, N. Guarino, C. Masolo, A. Oltramari, and L. Schneider, ‘Sweetening
Ontologies with DOLCE’, in A. Gómez-Pérez and V. R. Benjamins (eds.), Knowledge Engineering and Knowledge Management. Proceedings of the 13th International Conference (EKAW 2002), Berlin: Springer, 2002, pp. 166–181.
[25] J. C. Glenn, T. J. Gordon, and E. Florescu, 2012 State of the Future. Executive
Edition, Washington (DC), The Millennium Project, 2012.
[26] A. Gómez-Pérez, M. Fernández-López, and O. Corcho, Ontological Engineering,
with Examples from the Areas of Knowledge Management, e-Commerce and the
Semantic Web, London, Springer, 2004.
[27] N. Goodman, ‘A World of Individuals’, In J. M. Bochenski, A. Church, and N.
Goodman, The Problem of Universals: A Symposium, Notre Dame (IN), University
of Notre Dame Press, 1956, pp. 13–31.
[28] T. R. Gruber, ‘A Translation Approach to Portable Ontologies’, Knowledge Acquisition, 5 (1993), 199–220.
17
[29] T. R. Gruber, ‘Toward Principles for the Design of Ontologies Used for Knowledge
Sharing’, in [41], pp. 1–17; revised version in [42], pp. 907–928.
[30] T. R. Gruber, ‘Ontology’, in L. Liu and M. T. Özsu (eds.) Encyclopedia of Database
Systems, Berlin, Springer, 2009, pp. 1963–1965.
[31] N. Guarino, ‘The Ontological Level’, in R. Casati, B. Smith, and G. White (eds.),
Philosophy and the Cognitive Sciences. Proceedings of the 16th International
Wittgenstein Symposium, Vienna, Hölder-Pichler-Tempsky, 1994, pp. 443–456.
[32] N. Guarino, ‘Formal Ontology, Conceptual Analysis and Knowledge Representation’, in [42], pp. 625–640.
[33] N. Guarino, ‘Understanding, Building, and Using Ontologies’, International Journal of Human-Computer Studies, 46 (1997), 293–310.
[34] N. Guarino, ‘Formal Ontology in Information Systems’, in N. Guarino (ed.), Formal
Ontology in Information Systems. Proceedings of the 1st International Conference
(FOIS’98), Amsterdam, IOS Press, 1998, pp. 3–15.
[35] N. Guarino, ‘The Role of Identity Conditions in Ontology Design’, in C. Freksa and
D. M. Mark (eds.), Spatial Information Theory. Cognitive and Computational Foundations of Geographic Information Science (COSIT’99), Berlin, Springer, 1999,
pp. 221–234.
[36] N. Guarino, ‘The Ontological Level: Revisiting 30 Years of Knowledge Representation’, in A. T. Borgida, V. Chaudhri, P. Giorgini, and E. Yu (eds.), Conceptual Modeling: Foundations and Applications. Essays in Honor of John Mylopoulos, Berlin:
Springer, 2009, pp. 52–67.
[37] N. Guarino, ‘On the Semantics of Ongoing and Future Occurrence Identifiers’, in H.
C. Mayr, G. Guizzardi, and O. Pastor (eds.), Conceptual Modeling. Proceedings of
the 36th International Conference (ER 2017), Cham, Springer, 2017, pp. 477–490.
[38] N. Guarino and P. Giaretta, ‘Ontologies and Knowledge Bases: Towards a Terminological Clarification’, in N. J. I. Mars (ed.) Towards Very Large Knowledge
Bases: Knowledge Building and Knowledge Sharing, Amsterdam, IOS Press, 1995,
pp. 25–32.
[39] N. Guarino and G. Guizzardi, ‘In the Defense of Ontological Foundations for Conceptual Modeling’, Scandinavian Journal of Information Systems, 18 (2006), 115–
126.
[40] N. Guarino, D. Oberle, and S. Staab, ‘What Is an Ontology’, in the second edition
of [92], pp. 1–17.
[41] N. Guarino and R. Poli (eds.), International Workshop on Formal Ontology in Conceptual Analysis and Knowledge Representation, Padua, Ladseb-CNR, 1993.
[42] N. Guarino and R. Poli (eds.), The Role of Formal Ontology in the Information
Technology, special issue of International Journal of Human-Computer Studies, 43
(1995), 623–965.
[43] N. Guarino and C. Welty, ‘Identity, Unity, and Individuality: Towards a Formal
Toolkit for Ontological Analysis’, in W. Horn (ed.), Proceedings of the 14th European Conference on Artificial Intelligence (ECAI-2000), Amsterdam, IOS Press,
2000, pp. 219–223.
[44] N. Guarino and C. Welty, ‘A Formal Ontology of Properties’, in R. Dieng and
O. Corby (eds.), Knowledge Engineering and Knowledge Management: Methods,
Models, and Tools. Proceedings of the 12th International Conference (EKAW 2000),
Berlin, Springer, 2000, pp. 97–112.
18
[45] N. Guarino and C. Welty, ‘Identity and Subsumption’, in R. Green, C. A. Bean, and
S. H. Myaeng (eds.), The Semantics of Relationships: An Interdisciplinary Perspective, Dordrecht, Kluwer, 2002, pp. 111–126.
[46] N. Guarino and C. Welty, ‘Evaluating Ontological Decisions with OntoClean’,
Communications of the ACM, 45/2 (2002), 61–65.
[47] N. Guarino and C. Welty, ‘An Overview of OntoClean’, in [92], pp. 151–172.
[48] G. Guizzardi, N. Guarino, and J. P. A. Almeida, ‘Ontological Considerations about
the Representation of Events and Endurants in Business Models’, in M. La Rosa, P.
Loos, and O. Pastor (eds.), Business Process Management. Proceedings of the 14th
International Conference (BPM 2016), Cham, Springer, 2016, pp. 20–36.
[49] G. Guizzardi and G. Wagner, ‘A Unified Foundational Ontology and some Applications of it in Business Modeling’, in J. Grundspenkis and M. Kirikova (eds.), 16th
Conference on Advanced Information Systems Engineering (CAiSE’04). Workshops
Proceedings, Riga, Riga Technical University, 2004, vol. 3, pp. 129–143.
[50] P. Hayes, ‘The Naive Physics Manifesto’, in D. Michie (ed.), Expert Systems in the
Micro-electronic Age, Edinburgh, Edinburgh University Press, 1979, pp. 242–270.
[51] M. Heidegger, Vom Wesen der Wahrheit. Zu Platons Höhlengleichnis und Theätet,
Wintersemester 1931/32 (ed. H. Mörchen), Frankfurt, Klostermann, 1988; Eng.
trans. by T. Sadler, The Essence of Truth: On Plato’s Cave Allegory and Theaetetus,
London, Continuum, 2002.
[52] H. Herre, B. Heller, P. Burek, R. Hoehndorf, F. Loebe, and H. Michalek, ‘General
Formal Ontology (GFO). A Foundational Ontology Integrating Objects and Processes’, Technical Report, University of Leipzig, Research Group Ontologies in
Medicine (Onto-Med).
[53] J. R. Hobbs and R. C. Moore (eds.), Formal Theories of the Commonsense World,
Norwood (NJ), Ablex, 1985.
[54] E. Husserl, Logische Untersuchungen, Niemeyer, Halle, 1900–1901; 2nd edition,
1913 (vol. 1 and vol. 2.i) and 1921 (vol. 2.ii); Eng. trans. by J. N. Findlay, Logical
Investigations, London, Routledge and Kegan Paul, 1970.
[55] R. Ingarden, Der Streit um die Existenz der Welt, Tübingen, Niemeyer, 1964; partial
Eng. trans. by H. R. Michejda, Time and Modes of Being, Springfield (IL), Thomas,
1964.
[56] G. Jakus, V. Milutinovic, S. Omerović, and S. Tomažič, Concepts, Ontologies, and
Knowledge Representation, New York, Springer, 2013
[57] P. S. Kitcher, ‘Species’, Philosophy of Science, 51 (1984), 308–333.
[58] Y. Kompatsiaris and P. Hobson (eds.), Semantic Multimedia and Ontologies. Theory
and Applications, London, Springer, 2008.
[59] W. Kuśnierczyk, ‘Nontological Engineering’, in B. Bennett and C. Fellbaum (eds),
Formal Ontology in Information Systems. Proceedings of the 4th International Conference (FOIS-2006) Amsterdam, IOS Press, 2006, pp. 39–50.
[60] D. B. Lenat, ‘Cyc: A Large-Scale Investment in Knowledge Infrastructure’, Communications of the ACM, 38/11 (1995), 33–38.
[61] D. B. Lenat and R. V. Guha, Building Large Knowledge-Based Systems: Representation and Inference in the Cyc Project, Reading (MA), Addison-Wesley, 1990.
[62] D. K. Lewis, ‘Against Structural Universals’, Australasian Journal of Philosophy,
64 (1986), 25–46.
19
[63] D. K. Lewis, Parts of Classes, Oxford, Blackwell, 1991.
[64] E. H. Y. Lim, J. N. K. Liu, and R. S. T. Lee, Knowledge Seeker: Ontology Modelling
for Information Search and Management. A Compendium, Berlin, Springer, 2011.
[65] C. Masolo, S. Borgo, A. Gangemi, N. Guarino, and A. Oltramari, Wonderweb deliverable D18 (Technical Report), Trento, Laboratory for Applied Ontology ISTCCNR, 2003.
[66] C. Masolo, L. Vieu, E. Bottazzi, C. Catenacci, R. Ferrario, A. Gangemi, and N.
Guarino, ‘Social Roles and their Descriptions’, in D. Dubois, C. Welty, and M.-A.
Williams (eds.), Principles of Knowledge Representation and Reasoning. Proceedings of the 9th International Conference (KR2004), Menlo Park (CA), AAAI Press,
2004, pp. 267–277.
[67] D. McDermott, review of [61], Artificial Intelligence, 61 (1993), 53–63.
[68] A. Meinong, ‘Über Gegenstandstheorie’, in A. Meinong, R. Ameseder, and E.
Mally (eds.), Untersuchungen zur Gegenstandstheorie und Psychologie, Leipzig,
Barth, 1904, pp. 1–50; Eng. trans. by I. Levi, D. B. Terrell, and R. M. Chisholm,
‘On the Theory of Objects’, in R. M. Chisholm (ed.), Realism and the Background
of Phenomenology, Glencoe (IL), Free Press, 1960, pp. 76–117.
[69] G. H. Merrill, ‘Ontological Realism: Methodology or Misdirection?’, Applied Ontology, 5 (2010), 79–108.
[70] G. H. Merrill, ‘Realism and Reference Ontologies: Considerations, Reflections, and
Problems’, Applied Ontology, 5 (2010), 189–221.
[71] R. Mizoguchi,‘YAMATO: Yet Another More Advanced Top-level Ontology’, in K.
Taylor, T. Meyer, and M. Orgun (eds.), Advances in Ontologies. Proceedings of the
6th Australasian Ontology Workshop, Sydney, ACS, 2010, pp. 1–16.
[72] K. Munn and B. Smith (eds.), Applied Ontology: An Introduction, Frankfurt, Ontos,
2008.
[73] R. Neches, R. R. Fikes, T. Finin, T. R. Gruber, T. Senator, and W. R. Swartout,
‘Enabling Technology for Knowledge Sharing’, AI Magazine 12/3 (1991): 36–56.
[74] I. Niles and A. Pease, ‘Towards a Standard Upper Ontology,’ in C. Welty and B.
Smith (eds.), Formal Ontology in Information Systems. Proceedings of the 2nd International Conference (FOIS’01), New York, ACM Press, 2001, pp. 2–9.
[75] C. Partridge, Business Objects: Re-Engineering for Re-Use, Oxford, ButterworthHeinemann, 1996.
[76] R. Poli, M. Healy, and A. Kameas (eds.), Theory and Applications of Ontology:
Computer Applications, Berlin, Springer, 2010.
[77] H. Putnam, ‘The Meaning of ‘Meaning”, Minnesota Studies in Philosophy of Science, 7 (1975), 131–193.
[78] W. V. O. Quine, ‘Designation and Existence’, Journal of Philosophy, 36 (1939),
701–709.
[79] W. V. O. Quine, ‘On What There Is’, Review of Metaphysics, 2 (1948), 21–38.
[80] W. V. O. Quine, ‘Speaking of Objects’, Proceedings and Addresses of the American
Philosophical Association, 31 (1958), 5–22.
[81] W. V. O. Quine, Philosophy of Logic, Englewood Cliffs (NJ), Prentice-Hall, 1970.
[82] W. Rapaport, ‘Logical Foundations for Belief Representation’, Cognitive Science,
10 (1986), 371–422.
20
[83] E. Schrödinger, Science and Humanism. Physics in Our Time, Cambridge, Cambridge University Press, 1952.
[84] J. R. Searle, The Construction of Social Reality, New York, Free Press, 1995.
[85] R. Sharman, R. Kishore, and R. Ramesh (eds.), Ontologies. A Handbook of Principles, Concepts and Applications in Information Systems, Berlin, Springer, 2007.
[86] P. M. Simons, Parts. A Study In Ontology, Oxford, Clarendon Press, 1987.
[87] B. Smith, ‘Basic Concepts of Formal Ontology’, in N. Guarino (ed.), Formal Ontology in Information Systems. Proceedings of the 1st International Conference
(FOIS’98), Amsterdam, IOS Press, 1998, pp. 19–28.
[88] B. Smith, ‘Ontology’, in L. Floridi (ed.), The Blackwell Guide to the Philosophy of
Computing and Information, Oxford, Blackwell, 2003, pp. 155–166.
[89] B. Smith, ‘Beyond Concepts: Ontology as Reality Representation’, in A. C. Varzi
and L. Vieu (eds.), Formal Ontology in Information Systems. Proceedings of the 3rd
International Conference (FOIS-2004), Amsterdam, IOS Press, 2004, pp. 73–84.
[90] B. Smith and W. Ceusters, ‘Ontological Realism: A Methodology for Coordinated
Evolution of Scientific Ontologies’, Applied Ontology, 5 (2010), 139–188.
[91] B. Smith and C. Welty, ‘Ontology: Towards a New Synthesis’, in C. Welty and
B. Smith (eds.), Formal Ontology in Information Systems. Proceedings of the 2nd
International Conference (FOIS’01), New York, ACM Press, 2001, pp. iii–ix.
[92] S. Staab and R. Studer (eds.), Handbook on Ontologies, Berlin, Springer, 2004;
second edition: 2009.
[93] M. C. Suárez-Figueroa, A. Gómez-Pérez, E. Motta, and A. Gangemi (eds.), Ontology Engineering in a Networked World, Berlin, Springer, 2012.
[94] J. J. Thomson, ‘The Statue and the Clay’, Noûs, 32 (1998), 149–173.
[95] A. C. Varzi, ‘From Language to Ontology: Beware of the Traps’, in M. Aurnague,
M. Hickmann, and L. Vieu (eds.), The Categorization of Spatial Entities in Language and Cognition, Amsterdam, Benjamins, 2007, pp. 269–284.
[96] A. C. Varzi, ‘On the Boundary between Material and Formal Ontology’, in B.
Smith, R. Mizoguchi, and S. Nakagawa (eds.), Interdisciplinary Ontology, Vol. 3:
Proceedings of the Third Interdisciplinary Ontology Meeting, Tokyo, Keio University Press, 2010, pp. 3–8.
[97] A. C. Varzi, ‘On Doing Ontology without Metaphysics’, Philosophical Perspectives, 25 (2011), 407–423.
[98] A. C. Varzi, ‘Logic, Ontological Neutrality, and the Law of Non-Contradiction’, in
E. Ficara (ed.), Contradictions. Logic, History, Actuality, Berlin, De Gruyter, 2014,
pp. 53–80.
[99] C. Welty and N. Guarino, ‘Supporting Ontological Analysis of Taxonomic Relationships’, Data & Knowledge Engineering, 39 (2001), 51–74.
21