{ "paper_id": "Y07-1003", "header": { "generated_with": "S2ORC 1.0.0", "date_generated": "2023-01-19T13:46:18.726951Z" }, "title": "Deep Lexical Semantics: The Ontological Ascent *", "authors": [ { "first": "Jerry", "middle": [ "R" ], "last": "Hobbs", "suffix": "", "affiliation": { "laboratory": "", "institution": "University of Southern California Marina del Rey", "location": { "region": "California", "country": "USA" } }, "email": "hobbs@isi.edu" } ], "year": "", "venue": null, "identifiers": {}, "abstract": "Concepts of greater and greater complexity can be constructed by building systems of entities, by relating other entities to that system with a figure-ground relation, by embedding concepts of figure-ground in the concept of change, by embedding that in causality, and by coarsening the granularity and beginning the process over again. This process can be called the Ontological Ascent. It pervades natural language discourse, and suggests that to do lexical semantics properly, we must carefully axiomatize abstract theories of systems of entities, the figure-ground relation, change, causality, and granularity. In this paper, I outline what these theories should look like.", "pdf_parse": { "paper_id": "Y07-1003", "_pdf_hash": "", "abstract": [ { "text": "Concepts of greater and greater complexity can be constructed by building systems of entities, by relating other entities to that system with a figure-ground relation, by embedding concepts of figure-ground in the concept of change, by embedding that in causality, and by coarsening the granularity and beginning the process over again. This process can be called the Ontological Ascent. It pervades natural language discourse, and suggests that to do lexical semantics properly, we must carefully axiomatize abstract theories of systems of entities, the figure-ground relation, change, causality, and granularity. In this paper, I outline what these theories should look like.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Abstract", "sec_num": null } ], "body_text": [ { "text": "If we are going to have programs that understand language, we will have to encode what words mean. Since words refer to the world, their definitions will have to be in terms of some underlying theory of the world. We will therefore have to construct that theory, and do so in a way that reflects the ontology that is implicit in natural language.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "There are wrong ways to go about this enterprise. For example, we could take our underlying theory to be quantum mechanics and attempt to define, say, verbs of motion in terms of the primitives provided by that theory. A less obviously wrong approach, and one that has sometimes been tried, is to adopt Euclidean 3-space as the underlying model of space and attempt to define, say, spatial prepostions in terms of that.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "In this paper, I propose a general structure for a different underlying conceptualization of the world-one that should be particularly well suited to language. It consists of a set of core theories of a very abstract character. These theories are too abstract to impose many constraints on the entities and situations they are applied to. In fact, the reader may complain that they apply to anything. But the main purpose of the core theories is to provide the basis for a rich vocabulary for talking about entities and situations. The fact that the core theories apply so widely means that they provide a great many domains of discourse with a rich vocabulary.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "These will not be just any core theories, but theories that find their place in a schema I call the \"Ontological Ascent\". They include a theory of systems or composite entities and the figureground relation, which subsumes a theory of scales; a theory of change of state; a theory of causality, which provides support for a theory of goal-directed behavior; and very importantly a theory of shifts in granularity. The Ontological Ascent is described in Section 2. Then in the subsequent sections I sketch the outlines of each of the above theories.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "The enterprise is therefore to axiomatize these core theories in as clean a fashion as possible, and then to define, or at least characterize, various words in terms of predicates supplied by these core theories. For example, a core theory of scales will provide axioms involving predicates such as scale, <, subscale, top, bottom, and at. Then, at the \"lexical periphery\" we will be able to define the rather complex word \"range\" by an axiom such as the following:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "(\u2200 x, y, z)range(x, y, z) \u2261 (\u2203 s, s 1 , u 1 , u 2 )scale(s) \u2227 subscale(s 1 , s) \u2227 bottom(y, s 1 ) \u2227 top(z, s 1 ) \u2227 u 1 \u2208 x \u2227 at(u 1 , y) \u2227 u 2 \u2208 x \u2227 at(u 2 , z) \u2227 (\u2200 u \u2208 x)(\u2203 v \u2208 s 1 )at (u, v) That is, x ranges from y to z if and only if there is a scale s with a subscale s 1 whose bottom is y and whose top is z, such that some member u 1 of x is at y, some member u 2 of x is at z, and every member u of x is at some point v in s 1 . Many things can be conceptualized as scales, and when this is done, a large vocabulary, including the word \"range\", becomes available. Two methodological principles should be mentioned first. Above, I said \"define, or at least characterize, various words\". In general, we cannot hope to find definitions for words. That is, for very few words p will we find necessary and sufficient conditions, giving us axioms of the sort", "cite_spans": [ { "start": 187, "end": 193, "text": "(u, v)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "(\u2200 x)p(x) \u2261 . . .", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "Rather, we will find many necessary conditions and many sufficient conditions.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "(\u2200 x)p(x) \u2283 . . . (\u2200 x) . . . \u2283 p(x)", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "However, the accumulation of enough such axioms will tightly constrain the possible interpretations of the predicate, and hence the meaning of the word.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "The second methodological point is that we need to be careful how we use an argument from the \"naturalness\" of an expression. Not all expressions that will be allowed by our core theories will sound natural. Our knowledge of language consists of thousands of very specific conventions, each of which has a rationale in terms of core theories. But not everything that has a rationale has been conventionalized. Conventional expressions sound natural. Other expressions with a rationale are interpretable, but may not sound natural. For example, it is conventional to say \"at work\" and \"in progress\", and recently in corporate America, the expression \"on travel\" has become conventional. There is no particular reason that these expressions are better than \"on work\", \"on progress\", and \"at travel\". It just happens that the latter did not become conventional. The account of lexical meaning given here is intended to provide a rationale for expressions, but not to explain why one version rather than another has been conventionalized.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1." }, { "text": "A general pattern in natural language, illustrated in Figure 1 , consists of the following five steps:", "cite_spans": [], "ref_spans": [ { "start": 54, "end": 62, "text": "Figure 1", "ref_id": null } ], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "1. A collection of entities and relations among them can be viewed as a single system of entities.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "2. An element external to a system can be located at an element in a system.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "3. There can be a change of state from that element's being at one point to its being at another point in the system.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "4. An agent can cause such a change.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "5. At any point in this ascent, we can coarsen the granularity and view the complex entity that has been constructed as an indecomposable entity, itself ready to become the basis for another ascent. This ascent is reflected in the common pattern or decomposition for verbs, which we can represent, not quite formally, as cause (a, change(at(b, c) ", "cite_spans": [ { "start": 327, "end": 346, "text": "(a, change(at(b, c)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": ", at(b, d)))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "That is, a causes there to be a change of state from b being at c to b being at d. This schema, or something similar to it, has been central in much work on lexical semantics, including Gruber (1965) , Jackendoff (1976) , Hobbs (1974) , Talmy (1983) , and Croft (1991) .", "cite_spans": [ { "start": 186, "end": 199, "text": "Gruber (1965)", "ref_id": "BIBREF5" }, { "start": 202, "end": 219, "text": "Jackendoff (1976)", "ref_id": "BIBREF10" }, { "start": 222, "end": 234, "text": "Hobbs (1974)", "ref_id": "BIBREF6" }, { "start": 237, "end": 249, "text": "Talmy (1983)", "ref_id": "BIBREF11" }, { "start": 256, "end": 268, "text": "Croft (1991)", "ref_id": "BIBREF1" } ], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "The transitive verb \"move\" illustrates this pattern. If A moves B from C to D, then A causes a change from B being at C to B being at D.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "To illustrate just how pervasive this pattern is in natural language discourse, let us consider a random sentence taken from the business news section of a newspaper.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "In a stunning reversal for one of Silicon Valley's fastest-growing companies, Media Vision Technology Inc. said Thursday it will report a sharp decline in sales and \"a substantial loss\" in the quarter ending March 31 -a jolt that cut its stock price in half. (San Jose Mercury News, March 25, 1994, p. 12E) We will work from the inside out.", "cite_spans": [ { "start": 259, "end": 306, "text": "(San Jose Mercury News, March 25, 1994, p. 12E)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "Consider the word \"sales\". Possession can be conceived of as an \"at\" relation, an entity being at a person within the social network of people, the figure-ground relation. A sale is a change of state from an entity being at one person to the entity being at another person, and a corresponding change in the location of money. When we form the plural \"sales\", we are first coarsening the granularity to view a sale as an indecomposable entity, and then we are forming the simplest sort of system, a set, out of a collection of them.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Decline in sales\": We first coarsen the granularity on the set of sales, to view it as an indecomposable entity. This entity can be measured in amounts of money. To measure something is to locate it at a point on a particular kind of scale. Here the scale is amount of money. A decline then is a change of state from the set of sales (in a particular time period) being at one point on this scale to the sales (in a more current time period) being at a lower point on the scale. \"Sharp decline in sales\": Now the decline itself is viewed as an entity and located at the high end of a scale that encodes the rate of change on a quantitative scale with respect to time.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"A substantial loss\": A loss is also a change of state in a downward direction on a scale for amounts of money. Like \"sharp\", \"substantial\" is a measure of that change.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"A sharp decline in sales and a substantial loss\": The conjunction takes two events, viewed as indecomposable entities, and forms the aggregate consisting of the two of them. \"In the quarter ending March 31\": This adverbial applied to the conjoined NP locates the aggregate on the time scale within a particular interval.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Report\": Cognition can be viewed in terms of a figure-ground, or location \"at\", metaphor. We have ideas in mind, events in memory, and so on. Communication, on this metaphor, is a change of state from a proposition being in one mind to the proposition being in another mind. Of course, this sort of change of location differs from change of phyiscal location, in that the proposition is still in mind of the originator of the message. A report involves just such a communication event. The eventuality described as \"a sharp decline in sales and 'a substantial loss' in the quarter ending March 31\" moves from being located in the minds of the company officials to the minds of the public. Moreover, this change is caused by the entity issuing the report. \"Will report\": \"Will\" places the reporting action on the time line at some time in the future. \"Media Vision Technology Inc. said . . . it will report\": Saying is another communication event, and it is caused by the sayer. The future occurrence of the reporting event moves from the minds of the corporate officials to the minds of the public, and Media Vision Technology causes this to occur.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Media Vision Technology Inc. said Thursday it will report\": The adverbial \"Thursday\" locates the saying action on the time scale.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Growing\": Companies can be located on a scale of sizes. The growth of a company is a upward change of state on that scale.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Fastest-growing companies\": This change is viewed as indecomposable, and it is located on a scale of \"fast-ness\", its rate of change with respect to time. The set of growing companies is constructed, and their rates of change are located on this scale. A subset of these companies located highest on this scale is selected.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Silicon Valley's fastest-growing companies\": The possessive indicates a kind of \"at\" relation between each company and Silicon Valley.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Reversal\": A reversal is a change of state from motion on a scale in one direction to motion in the opposite direction. \"Growing\" gives the initial state, \"decline\" the final state.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Stunning\": A person's cognitive state can be characterized by the complexity of the thought processes that are possible. This complexity constitutes a scale on which the person can be located. To be stunned is first of all a downward change of state on this scale. Moreover, the size of that change is large, i.e., can be located in the high region of a scale of such changes, and it is rapid, i.e., can be located in the high region of a scale measuring the rate of such changes.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Stock price\": This phrase locates stocks on a scale of prices. \"Cut its stock price in half\": A cut is, at least, a downward change of state on a scale, in this case, of prices; this is the aspect of \"cut\" that is accessed in the metaphor. Because the scale of prices supports arithmetic, \"in half\" makes sense.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"Jolt\": A jolt is a large and sudden change of state in velocity and/or direction. Again, the change is viewed as indecomposable, and located on scales of size and rapidity.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "\"A jolt that cut its stock price in half\": This change in velocity and direction caused the change of state in stock Nearly every morpheme in this sentence introduces one or more scales or other kind of system, the location of an entity on a scale or in a system, a change in this location, or a causal relation in which a change is a cause or an effect.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "Beyond single sentences as well, we see the influence of the Ontological Ascent. The structure of discourse arises out of coherence relations that obtain between successive segments of discourse. These coherence relations can be broadly classified into four principal groups. There are relations, such as parallelism, based on similarity, the sort of similarity that is the basis for forming sets of entities, the simplest sort of system. There are figure-ground relations, such as cases where the first segment provides a backdrop and the second segment describes the action that plays out against that backdrop. There are relations based on change of state, such as the relation that I have called the Occasion relation (Hobbs, 1985a) . And there are relations based on causality, such as explanation.", "cite_spans": [ { "start": 722, "end": 736, "text": "(Hobbs, 1985a)", "ref_id": "BIBREF7" } ], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "All of this strongly suggests that theories of these concepts must be at the very foundation of any knowledge base adequate for natural language understanding. In this paper, I outline the structure of such theories.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "The Ontological Ascent", "sec_num": "2." }, { "text": "A road can be viewed as a line, a surface, or a volume. When we are planning a trip, we view it as a line. When we are driving on it, we have to worry about our placement on it to the right or left, so we think of it as a surface. When we hit a pothole, it becomes a volume to us.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Granularity", "sec_num": "3." }, { "text": "This shifting of granularity is a general property of cognition. We are very good at adopting small, on-the-spot theories of situations that include just the aspects relevant to our current concerns. Notions of granularity must pervade the core theories we build. Many concepts are inherently granularity-dependent, and many other concepts provide us with means for imposing granularities on situations.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Granularity", "sec_num": "3." }, { "text": "A granularity is defined by an indistinguishability relation \u223c, or equivalently, a set covering. If the set covering is a partition, the indistinguishability relation is transitive. An example is when we are concerned only with the country a location is in and not any finer discrimination. Any two locations in, say, South Korea would be indistinguishable under this relation. If the set covering is comprised of overlapping sets, the indistinguishability relation is not transitive. An example is when we do not distinguish any two points lying within 1 cm of each other. When we view a road as a line, we are not distinguishing between two points that are at the same place along its length, even though they are, for example, in different lanes.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Granularity", "sec_num": "3." }, { "text": "Shifts in granularity are a way of turning complex reasoning problems into simpler problems, with fewer entities and more regular properties.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Granularity", "sec_num": "3." }, { "text": "Granularities are often determined by functionality. A topographical map of the Sierras will show a hiking trail that is one foot wide, and not show a boulder that is twenty feet in diameter.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Granularity", "sec_num": "3." }, { "text": "A system, or composite entity, is a set of entities, their properties, and the relations among them. The concept of system captures the minimal complexity something must have in order for it to have structure. It is hard to imagine something that cannot be conceptualized as a system. For this reason, a vocabulary for talking about systems will be broadly applicable.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "The elements of a system can themselves be viewed as systems, and this gives us a very common example of shifting granularities. It allows us to distinguish between the structure and the function of an entity. The function of an entity in a system is its relations to the other elements of the system, its environment, while the entity itself is viewed as indecomposable. The structure of the entity is revealed when we decompose it and view it as a system itself. We look at it at a finer granularity.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "An important question any time we can view an entity both functionally and structurally is how the functions of the entity are implemented in its structure. We need to spell out the structurefunction articulations.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "For example, a librarian might view a book as an indecomposable entity and be interested in its location in the library, its relationship to other books, to the bookshelves, and to the people who check the book out. This is a functional view of the book with respect to the library. We can also view it structurally by inquiring as to its parts, its content, and so on. In spelling out the structure-function articulations, we might say something about how its content determines its place in the library.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "A system can serve as the ground against which some external figure can be located or can move. A primitive predicate at expresses this relation. In at(x, y, s) s is a system, y is an element in the system, and x is an entity not in the system. It says that the figure x is at a point y in the system s which is the ground.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "The at relation plays primarily two roles in the knowledge base. First, it is involved in the \"decompositions\" of many lexical items. We saw this above in the definition of \"range\". There is a very rich vocabulary of terms for talking about the figure-ground relation. This means that whenever a relation in some domain can be viewed as an instance of the figure-ground relation, we acquire at a stroke a rich vocabulary for talking about that domain.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "This gives rise to the second role the at predicate plays in the knowledge base. A great many specific domains have relations that can be stipulated to be instances of the at relation. There must be a large number of axioms of the form", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "(\u2200 x, y, s)r(x, y) \u2227 y \u2208 s \u2283 at(x, y, s)", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "Such axioms constitute the source of spatial terminology and spatial metaphors. Some examples of at relations are A person at an object in a system of objects:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "John is at his desk.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "An object at a location in a coordinate system: The post office is at the corner of 34th Street and Eighth Avenue.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "In computer science, a variable at a value in a range of values: I goes from 1 to 100.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "A person's salary at a particular point on the money scale: John's salary reached $75,000 this year.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "A particularly important example of an at relation is predication itself. We can view a set of predicates as constituting a system, where the relations among the elements are the implication and mutual exclusivity relations. Axioms of the form", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "(\u2200 p, x, s)p \u2208 s \u2227 p(x) \u2283 at(x, p, s)", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "say that for a predicate in a system of predicates to be true of an entity is for the entity to be at that predicate in the system. This makes the rich vocabulary of spatial relationships available for predication.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "The following expressions, for example, tap into a system of predicates about human activities and states of consciousness: at work, at play, on travel, on drugs, . . .", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Systems and the Figure-Ground Relation", "sec_num": "4." }, { "text": "A very common and very useful kind of system is one in which the relations among the entities are an indistinguishability relation \u223c and a partial ordering <. We can call this a scale.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Scales", "sec_num": "5." }, { "text": "A core theory of scales will provide definitions for such concepts as a subscale, a total ordering, a scale being dense, the top and bottom of a scale, and the reverse of a scale. Allen's relations among time intervals (Allen and Kautz, 1985) are in fact relations among subscales and are straightforward to define. If we have a primitive notion of points on a scale being adjacent, we can define connectedness in terms of it. A scale is a system, so the figure-ground relation applies to it. We can talk about an external entity being at a point on a scale.", "cite_spans": [ { "start": 219, "end": 242, "text": "(Allen and Kautz, 1985)", "ref_id": "BIBREF0" } ], "ref_spans": [], "eq_spans": [], "section": "Scales", "sec_num": "5." }, { "text": "The integers and the real numbers are scales, and many useful scales have structures isomorphic to these. At the other extreme, a scale might have only its two endpoints, as for the scale consisting of a predicate and its negation, {\u00acp, p}. But various intermediate levels of structure are often useful. In qualitative physics (de Kleer and Brown, 1985; Forbus, 1988) , many scales are reduced to negative, zero, and postive regions. It is often necessary to know not the precise value of a quantity but only its order of magnitude or half order of magnitude.", "cite_spans": [ { "start": 331, "end": 353, "text": "Kleer and Brown, 1985;", "ref_id": "BIBREF2" }, { "start": 354, "end": 367, "text": "Forbus, 1988)", "ref_id": "BIBREF4" } ], "ref_spans": [], "eq_spans": [], "section": "Scales", "sec_num": "5." }, { "text": "One important example of intermediate structure involves isolating the high and low regions of a scale. We can do this with operators we can call Hi and Lo. The Hi region of a scale includes its top; the Lo region includes its bottom. The points in the Hi region are all greater than any of the points in the Lo region. Otherwise, there are no general constraints on the Hi and Lo regions. In particular, the bottom of the Hi region and the top of the Lo region may be indeterminate with respect to the elements of the scale. The Hi and Lo operators provide us with a fairly coarse-grained structure on scales, useful when greater precision is not necessary or not possible.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Scales", "sec_num": "5." }, { "text": "The absolute form of adjectives frequently isolate Hi and Lo regions of scales. A totally ordered Height Scale can be defined precisely, but frequently we are only interested in qualitative judgments of height. The word \"tall\" isolates the Hi region of the Height Scale; the word \"short\" isolates the Lo region. A Happiness Scale cannot be defined precisely. We cannot get much more structure for a Happiness Scale than what is given to us by the Hi and Lo operators. The Hi and Lo operators can be iterated, to give us the concepts \"happy\", \"very happy\", and so on.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Scales", "sec_num": "5." }, { "text": "In any given context, the Hi and Lo operators will identify different regions of the scale. That is, the inferences we can draw from the fact that something is in the Hi region of a scale are context-dependent; indeed, inferences are always context-dependent. The Hi and Lo regions must be related to common distributions of objects on the scale, so that if something is significantly above average for the relevant set, then it is in the Hi region. They must also be related to goaldirected behavior; often something is in the Hi region of a scale precisely because that property aids or defeats the achievement of some goal in a plan. For example, saying that a talk is long often means that it is longer than the audience's attention span, and thus the goal of conveying information is defeated.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Scales", "sec_num": "5." }, { "text": "It is possible to define composite scales. If a scale s is a composite of scales s 1 and s 2 , then its elements are the ordered pairs < x, y > where x is in s 1 and y is in s 2 . An external entity is at a point < x, y > in the composite scale s if and only if it is at x in component scale s 1 and at y in component scale s 2 . The ordering in s has to be consistent with the orderings in s 1 and s 2 ; if x 1 is less than x 2 in s 1 , and y 1 is less than y 2 in s 2 , then < x 1 , y 1 > is less than < x 2 , y 2 > in s. The converse is not necessarily true; the composite scale may have more structure than that inherited from its component scales.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Scales", "sec_num": "5." }, { "text": "We need composite scales to deal with complex scalar predicates, such as damage. When something is damaged, it no longer fulfills its function in a goal-directed system. It needs to be repaired, and repairs cost. Thus, there are (at least) two ways in which damage can be serious, first in the degradation of its function, second in the cost of its repair. These are independent scales. Damage that causes a car not to run may cost next to nothing to fix, and damage that only causes the car to run a little unevenly may be very expensive.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Scales", "sec_num": "5." }, { "text": "The next primitive concept of central importance in the Ontological Ascent can be represented by the predicate change. This is a relation between situations, or conditions, or predications, or eventualities, and indicates a change of state. In this paper, to avoid an overgrowth of notation, I will write", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "change(p(x), q(x))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "where, strictly speaking, I should, in the ontologically promiscuous notation of Hobbs (1985a) , write change(e 1 , e 2 ) \u2227 p (e 1 , x) \u2227 q (e 2 , x)", "cite_spans": [ { "start": 81, "end": 94, "text": "Hobbs (1985a)", "ref_id": "BIBREF7" } ], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "This says that there is a change from the situation of p being true of x to the situation of q being true of x. A very common pattern involves a change of location:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "change(at(x, y, s), at(x, z, s))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "That is, there is a change from the situation of x being at y in s to x being at z in s. Here, at is the abstract figure-ground relation, so any domain conceptualized in terms of that automatically inherits the vocabulary provided by a theory of change.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "When there is a change, generally there is some entity involved in both the start and end states; there is something that is changing-x in the above formulas. This suggests a view of the world as consisting of a large number of more or less independent, occasionally interacting processes, or histories, or sequences of events. x goes through a series of changes, and y goes through a series of changes, and occasionally there is a state that involves a relation between the two.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "The predicate change possesses a limited transitivity. There was a change from Reagan being an actor to Reagan being President, because they are two parts of the same ongoing process, even though he was governor in between. But we probably do not want to say there was a change from Reagan being an actor to Margaret Thatcher being Prime Minister. They are not part of the same process.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "Any given process, that is, any sequence of events linked by change relations, is a scale whose partial ordering is induced by the predicate change.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "The Time Line could be taken as primitive, with the before relation as its ordering and an at-time relation relating states and events to points and intervals on the Time Line. The attime relation would be an at relation, giving us the common spatial metaphors for time. Such an ontology seems to be justified by the clock and calendar terms in modern languages. In this ontology, we could define the predicate change to be true when different properties are true of an entity at different times.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "It seems to me, however, that the notion of change is more basic. It is built into the more \"primitive\" parts of language, such as the event verbs. Even the words \"before\" and \"after\", which might seem to relate directly to the Time Line, carry a whiff of causality. The sentence,", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "The French Revolution broke out after George Washington was elected president. seems to convey some causality or suggest that somehow the two events are part of the same process.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Change", "sec_num": "6." }, { "text": "The next primitive predicate of central importance is cause. As with at and change, it has no definition. There is no axiom of the form (\u2200 e 1 , e 2 )cause(e 1 , e 2 ) \u2261 . . . but the knowledge base is rife with axioms of the form", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "cause(p(x), q(x))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "expressing causal connections among states and events. We don't know precisely what causality is, but we know lots and lots of examples of things that cause other things.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "There is a question as to what the arguments of cause can be. Some would urge that they can only be events, but it seems to me that we want to allow states as well, since in The slipperiness of the ice caused John to fall. the cause (the first argument) is a state. Moreover, intentional agents are sometimes taken to be the unanalyzed causes of events. In John lifted his arm.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "John is the cause of the change of position of his arm, and we probably don't want to have to coerce this argument into some imagined event taking place inside John. Physical forces may also act as causes, as in Gravity causes the moon to circle the earth.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "I have spoken loosely of states and events. We are now in a position to characterize more precisely the intuitive notions of state, event, action, and process. A state is an at relationship, at(x, y, s), or more generally, a predication. To be up, for example, is a state. An event is a change of state, a common variety of which is a change of location:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "change(at(x, y, s), at(x, z, s))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "For example, the verb \"rise\" denotes a change of location of something to a higher point. An action is the causing of an event by an intentional agent: cause(a, change(at(x, y, s), at(x, z, s)))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "The verb \"raise\" denotes an action by someone of effecting a change of location of something to a higher point. A process is a sequence of events or actions; in terms of the Ontological Ascent, we view particular events and actions as indecomposable and construct a system of them. For example, to fluctuate is to undergo a sequence of risings and fallings, and to pump is to engage in a sequence of raisings and lowerings. We can coarsen the granularity on processes so that the individual changes of state become invisible, and the result is a state. This is a transformation of perspective that is effected by the progressive aspect in English. Thus, fluctuating can be viewed as a state. State, event, action, and process label stages in the Ontological Ascent.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "The world is laced with threads of causal connection, and therefore our core theories must be rife with axioms encoding causal connections. In general, if two entities x and y are causally connected with respect to some behavior p of x, then whenever p happens to x, there is some corresponding behavior q that happens to y. Attachment of physical objects is one variety of causal connection. In this case, p and q are both move. If x and y are attached, moving x causes y to move. Containment is similar.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "A particularly common variety of causal connection between two entities is one mediated by the motion of a third entity from one to the other. cause(p(x), move(z, x, y)) \u2227 cause(move(z, x, y), q(y)) \u2283 cause(p(x), q(y))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "This might be called, somewhat facetiously, a \"vector boson\" connection. In particle physics, a vector boson is an elementary particle that transfers energy from one point to another. Photons, which really are vector bosons, mediate the causal connection between the sun and our eyes. Other examples of such causal connections are rain drops connecting a state of the clouds with the wetness of our skin and clothes, a virus transmitting disease from one person to another, and utterances passing information between people.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "Containment, barriers, openings, and penetration are all with respect to paths of causal connection.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "The event structure underlying many verbs exhibits causal chains. Instruments, for example, are usually vector bosons. In the sentence, John pounded the nail with a hammer for Bill.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "the underlying causal structure is that the Agent John causes a change in location of the Instrument, the hammer, which causes a change in location of the Object, the nail, which causes or should cause a change in the mental or emotional state of the Beneficiary, Bill.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "Agent -cause-> change(at(Instrument, x), at(Instrument, Object))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "-cause-> change(at(Object, y 1 ), at(Object, y 2 )) -cause-> change(p 1 (Benef iciary), p 2 (Benef iciary))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "Much of case grammar and work on thematic roles can be seen as a matter of identifying where the arguments of verbs fit into this kind of causal chain when we view the verbs as instantiating this abstract frame. An important role for causality is in linking two scales. It often effects a monotonic, scale-toscale function. The general pattern is this:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "cause(change(at(x, y, s 1 ), at(x, z, s 1 )), change(at(w, u, s 2 ), at(w, v, s 2 )))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "where if y < z on s 1 , then u < v on s 2 . That is, if there is a change from x being at y on s 1 to x being at a higher point z on s 1 , then this causes there to be a change from w being at u on s 2 to w being at a higher point v on s 2 . This is the basis of our many \"The more . . ., the more . . .\" rules, such as The more you press on the accellerator, the faster you go.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "The longer you leave the toast in, the browner it gets.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "We often conceive of graded causality in terms of the concept of \"force\". Talmy (1985) has shown how this concept pervades language.", "cite_spans": [ { "start": 74, "end": 86, "text": "Talmy (1985)", "ref_id": "BIBREF12" } ], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "A concept closely related to causality is enablement. It can be defined as follows:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "(\u2200 e 1 , e 2 )enable(e 1 , e 2 ) \u2261 cause(not(e 1 ), not(e 2 ))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "That is, e 1 enables e 2 if e 1 not happening will cause e 2 not to happen. Enablement is crucial in the core theory of goal-directed behavior.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Causality", "sec_num": "7." }, { "text": "A core theory of goal-directed behavior is central in human cognition and discourse and rests crucially on a core theory of causality. Its key primitive is the concept of a goal. This again will not be defined, but axioms will link it in the right way with axiomatizations of belief and action, to make available intentional interpretations of human and other behavior. In particular, these core theories should insure that people's actions can be seen as attempts to achieve their goals, given their beliefs. Among the most important facts about goals are those linking them with causality and enablement, for it is by manipulating the causal structure of the world that agents achieve their goals. The two primary axioms are as follows:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "(\u2200 a, q, r)goal(a, q) \u2227 enable(r, q) \u2283 goal(a, r) (\u2200 a, p, q)goal(a, q) \u2227 cause(p, q) \u2227 choose(a, p, q) \u2283 goal(a, p)", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "The first axiom says that if an agent a has a goal q and r enables q, then a will have the goal r. This captures the prerequisites of the STRIPS operators of Fikes and Nilsson (1971) .", "cite_spans": [ { "start": 158, "end": 182, "text": "Fikes and Nilsson (1971)", "ref_id": "BIBREF3" } ], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "The second axiom says that if an agent a has a goal q, where p causes q, and a chooses p as a way of achieving q, then a will have the goal p. I will not attempt to explicate choose here, but something like this is necessary to accomodate nondeterminism. There may be many things that will cause q, and the agent need pick only one of them. This axiom encodes the \"body\" of the STRIPS operators of Fikes and Nilsson (1971) .", "cite_spans": [ { "start": 398, "end": 422, "text": "Fikes and Nilsson (1971)", "ref_id": "BIBREF3" } ], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "Thus, to achieve a goal, an agent must satisfy all the prerequisites, removing all the barriers to the goal, and then choose something that will cause the goal to come about. These two axioms allow us to construct hierarchical plans, decomposing goals into their subgoals. In the above axioms, q is the goal, r and p the subgoals. These subgoals can in turn be decomposed into further subgoals", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "The depth of decomposition in these plans is one of the prinicpal ways we impose a granularity on our view of behavior. It may be sufficient for our purposes to know that John drove his car to the airport, or it may be necessary to view it under a finer granularity that makes visible his actions of shifting the gears and turning the steering wheel.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "A plan is essentially a representation of causal structure. It is therefore useful for explaining not just human behavior, but other phenomena as well. Artifacts and organizations can, for example, be viewed as plans made concrete.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "Much of the knowledge we have about an artifact is best represented by the plan that it implements. Consider a very simple example. The function of a coffee cup is to move coffee. We decompose this goal into two subgoals-containing the coffee in the cup and moving the cup. The subgoal of moving the cup is further decomposed into the subgoals of attaching the cup to the handle and moving the handle. It is a very common schema for artifacts that in order to do something to an object, we set up a causal connection, such as containment or attachment, to another object, and do something to that other object. We can continue to decompose in this fashion until we have specified the role or function of all the components of the artifact.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "Similarly, organizations can be seen as having a goal and implementing a plan to achieve that goal, where the structure of the organization reflects the structure of the plan. Thus, the goal of an organization might be to provide people with cars. This decomposes into the subgoals of having one division of the organization manufacture cars and another division sell them to people. Each of these would decompose further. Eventually the plan would bottom out in sets of actions by single individuals. These sets of actions constitute the members' roles in the organization.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "Any system that can be viewed as exhibiting functionality can be represented in terms of a plan that expresses the system's underlying causal structure. A tree, for example, can be viewed as a goal-directed system whose goal is to grow and reproduce.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Goals-Directed Behavior", "sec_num": "8." }, { "text": "A common way to encode knowledge for natural language and other AI programs is to proceed domain by domain. Many of the most common words in language, however, apply across many domains. What I have tried to do in this paper is to suggest some very abstract domains-systems and the figure-ground relation, scales, change of state, causality, and goal-directed systems-that seem to underlie more specific domains.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Summary", "sec_num": "9." }, { "text": "Moreover, concepts of greater and greater complexity can be constructed by building systems of entities, by relating other entities to that system with a figure-ground relation, by embedding concepts of figure-ground in the concept of change, by embedding that in causality, and by coarsening the granularity and beginning the process over again-a process I have called the Ontological Ascent.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Summary", "sec_num": "9." }, { "text": "More specific domains can be seen as instantiations of these abstract ones. Language provides us with a rich vocabulary for talking about the abstract domains. When we construct core theories of these domains, then we have a hope of being able to define, or at least characterize, the words in this vocabulary in terms provided by the core theories. When the core theory of an abstract domain is instantiated as a specific domain, then the vocabulary associated with the abstract domain is also instantiated, giving us a rich vocabulary for talking about the specific domain. Conversely, when we encounter general words in the contexts of specific domains, understanding how the specific domains instantiate the abstract domains allows us to determine the specific meanings of the general words in their current context.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Summary", "sec_num": "9." } ], "back_matter": [ { "text": "This paper is based in part on work that was done with William Croft, Todd Davies, Douglas Edwards, and Kenneth Laws. I have also profited from discussions with Annelise Bech, Andrew Gordon, Bente Maegaard, and Costanza Navarretta.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Acknowledgments", "sec_num": null } ], "bib_entries": { "BIBREF0": { "ref_id": "b0", "title": "A model of naive temporal reasoning", "authors": [ { "first": "James", "middle": [ "F" ], "last": "Allen", "suffix": "" }, { "first": "A", "middle": [], "last": "Henry", "suffix": "" }, { "first": "", "middle": [], "last": "Kautz", "suffix": "" } ], "year": 1985, "venue": "Formal Theories of the Commonsense World", "volume": "", "issue": "", "pages": "251--268", "other_ids": {}, "num": null, "urls": [], "raw_text": "Allen, James F. and Henry A. Kautz. 1985. A model of naive temporal reasoning. In Jerry R. Hobbs and Robert C. Moore, editors, Formal Theories of the Commonsense World. Ablex Publishing Corp., pages 251-268.", "links": null }, "BIBREF1": { "ref_id": "b1", "title": "Syntactic Categories and Grammatical Relations: The Cognitive Organization of Information", "authors": [ { "first": "William", "middle": [], "last": "Croft", "suffix": "" } ], "year": 1991, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Croft, William. 1991. Syntactic Categories and Grammatical Relations: The Cognitive Organi- zation of Information. University of Chicago Press, Chicago, Illinois.", "links": null }, "BIBREF2": { "ref_id": "b2", "title": "A qualitative physics based on confluences", "authors": [ { "first": "Johann", "middle": [], "last": "De Kleer", "suffix": "" }, { "first": "John", "middle": [], "last": "Seely Brown", "suffix": "" } ], "year": 1985, "venue": "Formal Theories of the Commonsense World", "volume": "", "issue": "", "pages": "109--184", "other_ids": {}, "num": null, "urls": [], "raw_text": "de Kleer, Johann and John Seely Brown. 1985. A qualitative physics based on confluences. In Jerry R. Hobbs and Robert C. Moore, editors, Formal Theories of the Commonsense World. Ablex Publishing Corp., pages 109-184.", "links": null }, "BIBREF3": { "ref_id": "b3", "title": "Strips: A new approach to the application of theorem proving to problem solving", "authors": [ { "first": "Richard", "middle": [], "last": "Fikes", "suffix": "" }, { "first": "Nils", "middle": [ "J" ], "last": "Nilsson", "suffix": "" } ], "year": 1971, "venue": "Artificial Intelligence", "volume": "2", "issue": "", "pages": "189--208", "other_ids": {}, "num": null, "urls": [], "raw_text": "Fikes, Richard and Nils J. Nilsson. 1971. Strips: A new approach to the application of theorem proving to problem solving. Artificial Intelligence, 2:189-208.", "links": null }, "BIBREF4": { "ref_id": "b4", "title": "Qualitative physics: Past, present, and future", "authors": [ { "first": "Kenneth", "middle": [ "D" ], "last": "Forbus", "suffix": "" } ], "year": 1988, "venue": "Readings in Qualitative Reasoning about Physical Systems", "volume": "", "issue": "", "pages": "11--39", "other_ids": {}, "num": null, "urls": [], "raw_text": "Forbus, Kenneth D. 1988. Qualitative physics: Past, present, and future. In Howard Shrobe, editor, Exploring Artificial Intelligence. Morgan Kaufmann Publishers, Inc., pages 239-296. Republished in Daniel S. Weld and Johan de Kleer (eds.), Readings in Qualitative Reasoning about Physical Systems, Morgan Kaufmann Publishers, Inc., San Mateo, California, 11-39, 1989.", "links": null }, "BIBREF5": { "ref_id": "b5", "title": "Studies in Lexical Relations", "authors": [ { "first": "Jeffrey", "middle": [ "C" ], "last": "Gruber", "suffix": "" } ], "year": 1965, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Gruber, Jeffrey C. 1965. Studies in Lexical Relations. Ph.D. thesis, Massachusetts Institute of Technology, Cambridge, Massachusetts. unpublished.", "links": null }, "BIBREF6": { "ref_id": "b6", "title": "A model for natural language semantics, part i: The model", "authors": [ { "first": "Jerry", "middle": [ "R" ], "last": "Hobbs", "suffix": "" } ], "year": 1974, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Hobbs, Jerry R. 1974. A model for natural language semantics, part i: The model. Technical Report 36, Department of Computer Science, Yale University, October. Research Report.", "links": null }, "BIBREF7": { "ref_id": "b7", "title": "On the coherence and structure of discourse", "authors": [ { "first": "Jerry", "middle": [ "R" ], "last": "Hobbs", "suffix": "" } ], "year": 1985, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Hobbs, Jerry R. 1985a. On the coherence and structure of discourse. Technical Report CSLI-85- 37, Center for the Study of Language and Information, Stanford University.", "links": null }, "BIBREF8": { "ref_id": "b8", "title": "Ontological promiscuity", "authors": [ { "first": "Jerry", "middle": [ "R" ], "last": "Hobbs", "suffix": "" } ], "year": 1985, "venue": "Proceedings, 23rd Annual Meeting of the Association for Computational Linguistics", "volume": "", "issue": "", "pages": "61--69", "other_ids": {}, "num": null, "urls": [], "raw_text": "Hobbs, Jerry R. 1985b. Ontological promiscuity. In Proceedings, 23rd Annual Meeting of the Association for Computational Linguistics, pages 61-69, Chicago, Illinois, July.", "links": null }, "BIBREF9": { "ref_id": "b9", "title": "Commonsense metaphysics and lexical semantics", "authors": [ { "first": "Jerry", "middle": [ "R" ], "last": "Hobbs", "suffix": "" }, { "first": "William", "middle": [], "last": "Croft", "suffix": "" }, { "first": "Todd", "middle": [], "last": "Davies", "suffix": "" }, { "first": "Douglas", "middle": [], "last": "Edwards", "suffix": "" }, { "first": "Kenneth", "middle": [], "last": "Laws", "suffix": "" } ], "year": 1987, "venue": "Computational Linguistics", "volume": "13", "issue": "3-4", "pages": "241--250", "other_ids": {}, "num": null, "urls": [], "raw_text": "Hobbs, Jerry R., William Croft, Todd Davies, Douglas Edwards, and Kenneth Laws. 1987. Com- monsense metaphysics and lexical semantics. Computational Linguistics, 13(3-4):241-250, July-December.", "links": null }, "BIBREF10": { "ref_id": "b10", "title": "Toward an explanatory semantic representation", "authors": [ { "first": "Ray", "middle": [], "last": "Jackendoff", "suffix": "" } ], "year": 1976, "venue": "Linguistic Inquiry", "volume": "7", "issue": "1", "pages": "89--150", "other_ids": {}, "num": null, "urls": [], "raw_text": "Jackendoff, Ray. 1976. Toward an explanatory semantic representation. Linguistic Inquiry, 7(1):89-150.", "links": null }, "BIBREF11": { "ref_id": "b11", "title": "How language structures space", "authors": [ { "first": "Leonard", "middle": [], "last": "Talmy", "suffix": "" } ], "year": 1983, "venue": "Spatial Orientation: Theory, Research, and Application", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Talmy, Leonard. 1983. How language structures space. In Herbert Pick and Linda Acredolo, editors, Spatial Orientation: Theory, Research, and Application. Plenum Press.", "links": null }, "BIBREF12": { "ref_id": "b12", "title": "Force dynamics in language and thought", "authors": [ { "first": "Leonard", "middle": [], "last": "Talmy", "suffix": "" } ], "year": 1985, "venue": "Proceedings from the Parasession on Causatives and Agentivity, 21st Regional Meeting", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Talmy, Leonard. 1985. Force dynamics in language and thought. In William H. Eilfort, Paul D. Kroeber, and Karen L. Peterson, editors, Proceedings from the Parasession on Causatives and Agentivity, 21st Regional Meeting, Chicago Linguistic Society.", "links": null } }, "ref_entries": { "FIGREF0": { "uris": null, "num": null, "text": "Figure 1: The Ontological Ascent", "type_str": "figure" } } } }