博文
Discuss on Ontology
|||
'ontology' has a very different sense in philosophy and in information science.
Representation of conceptual domains. This is the information science use, and is due to Tom Gruber, who was generally influenced by Husserlian formal ontology.
In analytic philosophy, an ontology is determined by an *interpretation*[解释，诠释], which is - loosely speaking - a function that projects an element of a language onto an element of the world, however widely-set or limited ‘the world’ might be. In more formal semantics, an interpretation determines a model by assigning values to the variables of a theory; if the theory is of first or higher order, then its domain will be those individuals over which the quantifiers range and the properties and relations which obtain between and among them.
So, in analytic metaphysics[形而上学], an ontology is usually determined by an interpretation of a first-order theory; and a first-order theory is (evidently) expressed in first-order logic. This, I think, provides a "unified language standard" that is sufficient to anyone's requirement ;-)
//提供统一的标准符合每个人的需求
Of course, this is entirely irrelevant to your particular practical interrogations[疑问]. But I surmise[猜测] that your puzzlement is due to the divergence[分歧] between:
1.*extensional ontology*, [本体外延]which is the ontology of first-order theories, and is derived from the Frege-Russell school through Carnap, the Polish logicians, and Quine. Such ontologies are concerned with the *extension* of concepts
//本体外延
2. *formal ontology*, which is derived from Brentano and Husserl; Barry Smith is a typical theorist in this field and any top-level ontology (BFO, DOLCE, GFO...) gives an example. Such ontologies are concerned with *concepts* and the relations between concepts
//形式化的本体，关注概念和概念与概念间的关系
3. *applied ontologies*, which derive from formal ontologies through Tom Gruber’s use of the term, are representations of a certain domain of knowledge. Vanrullen and Hirst (forthcoming) will show that "ontology" is, in fact, a misnomer here: as such ontologies concern concepts and the representation of knowledge, they range over *semantic[语义的]* and *epistemic[认知的]* domains, rather than *ontic* domains in sense (1) above; they would perhaps be better described as "applied epistemologies[应用认识论]".
//应用本体，对某领域知识的表述，
Applied ontologies in sense (3) range over domain-specific concepts and the relations obtaining between such concepts: thus, the ‘objects’ of such ontologies are not ‘regions of the world’, but ‘concepts’. The ontologies I describe (in sense (1)) range over the *extensions* of concepts - that is, things in the world that supposedly correspond to, exemplify, or satisfy concepts. [适用，演绎，满足概念]
As far as ‘social systems’ are concerned: //关注社会系统
Generally, the relation between a concept and the range or sequence of things in the world that might satisfy that concept is a matter of *determination* - of how well the concept picks out one specific thing or class of things in the world. Where an applied ontology models a well-defined taxonomy[分类] (as in ontologies applied to biological sciences) or a domain having a prescribed or semi-formalised language (as in air traffic control), the concepts which figure in the ontology have a clear, well-defined extension. Practical tools based on such ontologies (decision-making tools, diagnostic tools) function efficiently because there’s not much room for concept-based indetermination. [基于概念的不确定性]
However, when the ontology ranges over concepts with a less well-defined extension (as in most social interactions), the content of the concepts is not fixed and objectively-determinable in every situation, but is normative and context-sensitive. In everyday interaction, the precise sense of a concept depends on a context-sensitive and speaker-sensitive *interpretation* - and here, there’s a strong risk of underdetermination of any concept with respect to the range of things and situations ‘in the world’ to which it might apply. Such indetermination can touch terms with an apparently technical definition – and here we can take the different understandings of the term “ontology” as an example. It follows that indetermination is common where terms have social and cultural connotations that will be understood differently by different actors (this is a common problem in employer-employee relations, where notions of the nature and finalities of collective action can be understood very differently by operational agents, management, and direction); such distinctions are often determined by the sociolect of the interlocutors. The possibilities of indetermination are already large in exchanges between sociolects in one language and within a common general culture; they’re evidently far greater in exchanges between interlocutors having different native languages and no shared general culture.
All this being the case (and this rejoins Gustavo’s interrogations), any ‘tools’ which might apply to ‘social situations’ (sociometric[社会计量] tools, tools for managing and planning the disposition of human resources, tools for skills management, training tools etc.) need to integrate norm-based[基于规范] rather than rule-based[基于规则] conceptual domains (domains in which a term takes a range of possible values or sets of values that are further determined by context, rather than domains in which a concept takes one, fixed value or set of values).
A particular problem in the use of such devices as is to decision or consulting is that the user (and, for that matter, the technician) can confuse ‘concepts’ with the putative ‘extension of concepts’ – for example, that he or she can take for granted not only that “everyone knows what a family is”, but also that “everyone knows that a family is a system”.
The first is perhaps pragmatically the case for all competent speakers of English, but the second depends on the speaker having a reasonably sophisticated understanding of what a ‘system’ is... that is, on the speaker possessing a ‘theory of systems系统理论’. Indeed, this is precisely the ‘weakness’ of most systems thinking – systems theory originated with biologists, and biological entities (living organisms, organs, cells...) are probably the most “self-evident” class of systems in the natural world. 系统理论起源生物学，生物实体
However, while it might be evident to a biologist that something is a system, this is a product of the biologist’s particular epistemic context – it doesn’t follow that it is *universally* evident that that thing is a system. The biological systems theorist is confusing混淆 a concept – “organism”, “organ” – with the extension of that concept: a liver is a system *with respect to a system of theories* - but with respect to[关于] the man on the Clapham omnibus, it’s a piece of offal[废物] that he’s going to fry with onions for his tea.
This material came from talking with David Hirst..
https://wap.sciencenet.cn/blog-87352-392257.html
上一篇：tips for job hunting
下一篇：Predicate Logic
