As previously indicated, the ultimate goal of a general (or domain-independent) ontology is to provide a conceptual foundation valid for all possible domain ontologies. To attain this by now or in the near future may, originally, have appeared to be a reasonable expectation. Alas, it turned out that the obstacles are enormous. Despite considerable effort on the part of philosophers and information scientists, to create a top-level ontology that seamlessly relates to all possible domain ontologies still appears to be an elusive goal. The logical, semantic and syntactic complexities to attain such a goal turned out to be much greater than expected. Thus, the huge literature accumulated on this topic over the last two or three decades still constitutes a ‘promissory note’ for the future rather than cash on the table. There are several reasons for this relatively pessimistic view, but one of the major obstacles (beyond the ones mentioned above) is the disagreement among experts on the basic notions of a highest-level ontology. Again and again, one encounters reference to the ‘tower of Babel syndrome’. And this may well be the best analogy for the present state of ontology, particularly the new area of ‘systems ontology’. At a ﬁrst glance, our attempts (in creating a truly common ontological basis) appeared to have failed because of too much individualism and a great diversity of views. Hence, one might presume that we have to create a kind of ‘Esperanto’ of ontology where a large enough number of scholars can agree on a common basis – instead of each, or each group, launching its own conceptual and semantic apparatus. However, this approach too might fail, and for the same reason why Esperanto failed to become the lingua franca of a worldwide linguistic communication system. But successful lingua francas did exist in the past and still exist in the present: Latin, during the Middle Ages and English since the early 1950s. And the reason for this success is that both Latin and English are not artiﬁcial creations but living languages. Here, English has even the advantage over mediaeval Latin in so far as the latter was no longer really ‘alive’ among common people during the Middle Ages, though it was regularly enough used among priests, monks, physicians, and scholars to fulﬁl
the immensely important function of a lingua franca within the European intellectual community. But what is the lesson for ontology that we can draw from this analogy? I believe philosophers have to ﬁnd a common conceptual basis that no group of academics deems to be an imposition. In other words, we need a basis that is widely accepted as ‘natural’. But that is easier said than done. Thus it is hardly surprising if Peter Lyster of NIGMS (the National Institute of General Medical Science) points out that one of the most challenging and rewarding research areas is ‘the sociology of ontology building – how to get communities to develop and agree on standards’ (cf. Dutchen 2011: 2).1 Yet, if we should ever attain such a stage of enlightenment, it will be only by continuing efforts, scholarly scrutiny and experiments, practical tests as well as by trial and error. But others, like Stephanie Dutchen, are less optimistic; she believes that: ‘There probably will never be a single undisputed ontology that contains all scientiﬁc knowledge’ (ibid: 2.)
Let me give a few examples of the diversity of various attempts in constructing a basis for a general ontology. Apart from traditional philosophical attempts by Edmund Husserl (1900-01, 1913), Nicolai Hartmann (1933, 1940, 1948, 1953) and many other philosophers, one of the ﬁrst comprehensive ontologies for artiﬁcial intelligence (here referred to as ‘systems ontologies’) was the Cyc-project. It was conceived in 1984 by Douglas Lena and has since been developed by Cycorp that makes freely available parts of its results under the OpenCyc programme:
OpenCyc is the open source version of the Cyc technology, the world’s largest and most complete general knowledge base and commonsense reasoning engine. OpenCyc contains the full set of (non-proprietary) Cyc terms as well as millions of assertions. . . . Cycorp offers this ontology at no cost and encourages you to make use of it as you see ﬁt.