Anankastic conditionals are analyzed in terms of events conceived as sequences of snapshots – roughly, comics. Quantification is applied not to worlds (sets of which are customarily identified with propositions) but to strings that record observations of actions. The account generalizes to other types of conditionals, sidestepping certain well-known problems that beset possible worlds treatments, such as logical omniscience and irrelevance. A refinement for anankastic conditionals is considered, incorporating action relations.

A fundamental question in reasoning about change is, what information does a reasoning agent infer about later times from earlier times? I will argue that reasoning about change by an agent is to be modeled in terms of the persistence of the agent''s beliefs over time rather than the persistence of truth and that such persistence is explained by pragmatic factors about how agents acquire information from other agents rather than by general principles of persistence about states of the world. (...) AI accounts of persistence have focused on closed world examples of change, in which the agent believes that the truth of a proposition is unaltered so long as he or she has no evidence that it has been changed. AI principles of persistence seem plausible in a closed world where one assumes the agent knows everything that is happening. If one drops the assumption of omniscience, however, the analysis of persistence is implausible. To get a good account of persistence and reasoning about change, I argue we should examine open world examples of change, in which the agent is ignorant of some of the changes occurring in the world. In open world examples of change, persistence must be formulated, I argue, as a pragmatic principle about the persistence of beliefs. After elaborating my criticisms of current accounts of persistence, I examine how such pragmatic principles fare with the notorious examples of reasoning about action that have collectively characterized the so-called frame problem. (shrink)

This paper argues that chance (risk or opportunity) discovery is challenging, from a reasoning point of view, because it represents a dilemma for inductive reasoning. Chance discovery shares many features with the grue paradox. Consequently, Bayesian approaches represent a potential solution. The Bayesian solution evaluates alternative models generated using a temporal logic planner to manage the chance. Surprise indices are used in monitoring the conformity of the real world and the assessed probabilities. Game theoretic approaches are proposed to deal with (...) multi-agent interaction in chance management. (shrink)

The idea that natural language grammar and planned action are relatedsystems has been implicit in psychological theory for more than acentury. However, formal theories in the two domains have tendedto look very different. This article argues that both faculties sharethe formal character of applicative systems based on operationscorresponding to the same two combinatory operations, namely functional composition and type-raising. Viewing them in thisway suggests simpler and more cognitively plausible accounts of bothsystems, and suggests that the language faculty evolved in the (...) speciesand develops in children by a rather direct adaptation of a moreprimitive apparatus for planning purposive action in the world bycomposing affordances of objects or tools. Theknowledge representation that underlies such planning is alsoreflected in the natural language semantics of tense, mood, andaspect, which the paper begins by arguing provides the key tounderstanding both systems. (shrink)

Most agents can acquire information about their environments as they operate. A good plan for such an agent is one that not only achieves the goal, but is also executable, i.e., ensures that the agent has enough information at every step to know what to do next. In this paper, we present a formal account of what it means for an agent to know how to execute a plan and to be able to achieve a goal. Such a theory is (...) a prerequisite for producing specifications of planners for agents that can acquire information at run time. It is also essential to account for cooperation among agents. Our account is more general than previous proposals, correctly handles programs containing loops, and incorporates a solution to the frame problem. It can also be used to prove programs containing sensing actions correct. (shrink)

The paper proposes a semantics for contextual (i.e., Temporal and Locative) Prepositional Phrases (CPPs) like during every meeting, in the garden, when Harry met Sally and where I’m calling from. The semantics is embodied in a multi-modal extension of Combinatory Categoral Grammar (CCG). The grammar allows the strictly monotonic compositional derivation of multiple correct interpretations for “stacked” or multiple CPPs, including interpretations whose scope relations are not what would be expected on standard assumptions about surfacesyntactic command and monotonic derivation. A (...) type-hierarchy of functional modalities plays a crucial role in the specification of the fragment. (shrink)

We respond to Morris and Richardson 's claim that Pickering and Chater's arguments about the lack of a relation between cognitive science and folk psychology are flawed. We note that possible controversies about the appropriate uses for the two terms do not affect our arguments. We then address their claim that computational explanation of knowledge-rich processes has proved possible in the domains of problem solving, scientific discovery, and reasoning. We argue that, in all cases, computational explanation is only possible for (...) aspects of those processes that do not make reference to general knowledge. We conclude that consideration of the issues raised by Morris and Richardson reinforces our original claim that there are two fundamentally distinct projects for understanding the mind, one based on justification, and the other on computational explanation, and that these apply to non-overlapping aspects of mental life. (shrink)

It is often assumed that cognitive science is built upon folk psychology, and that challenges to folk psychology are therefore challenges to cognitive science itself. We argue that, in practice, cognitive science and folk psychology treat entirely non-overlapping domains: cognitive science considers aspects of mental life which do not depend on general knowledge, whereas folk psychology considers aspects of mental life which do depend on general knowledge. We back up our argument on theoretical grounds, and also illustrate the separation between (...) cognitive scientific and folk psychological phenomena in a number of cognitive domains. We consider the methodological and theoretical significance of our arguments for cognitive science research. (shrink)

Figure 1: A pr ototyp ical exa mple of a three-layer feed forward network, used by Plunkett and M archm an (1 991 ) to simulate learning the past-tense of En glish verbs. The inpu t units encode representations of the three phonemes of the present tense of the artificial words used in this simulation. Th e netwo rk is trained to produce a representation of the phonemes employed in the past tense form and the suffix (/d/, /ed/, or /t/) (...) used on regular verbs. To run the network, each input unit is assigned an activation value o f 0 or 1 , dep ending on whethe r the featu re is present or not. Eac h input unit is conne cted to each of the 30 hidden units by a we ighted conn ection and p rovid es an inp ut to each hidden unit equal to the product of the input unit's activation and the weight. Each hidd en unit's activation is then determined by summing ov er the va lues co ming fro m each inp ut unit to deter mine a netinput, and then applying a non-linear function (e.g., the logistic function 1/(1+e^netinput). Th is whole proced ure is. (shrink)

This paper continues a strain of intellectual complaint against the presumptions of certain kinds of formal semantics (the qualification is important) and their bad effects on those areas of artificial intelligence concerned with machine understanding of human language. After some discussion of the use of the term epistemology in artificial intelligence, the paper takes as a case study the various positions held by McDermott on these issues and concludes, reluctantly, that, although he has reversed himself on the issue, there was (...) no time at which he was right. (shrink)

The term attention has been used to mean so many different things that some have despaired of it being useful at all. This paper is devoted to bringing a modicum of order to the chaos through the time-honored device of categorization. The chief purpose of this paper is to introduce a comprehensive descriptive taxonomy of the nuanced ways the term attention may be employed. It is presented in table form, followed by elucidations and illustrations of each of its items. But (...) first, I offer reasons why a taxonomy of attention is needed and explore some of its possible benefits. After presenting the taxonomy, I sketch by way of example how it might be applied to two interesting questions: is the umbrella term attention still useful?; and, what is it that ties the subdefinitions of attention together yet distinguishes them from other kinds of cognition? (shrink)

Showing remarkable insight into the relationship between language and thought, Alan Turing in 1950 proposed the Imitation Game as a proxy for the question “Can machines think?” and its meaning and practicality have been debated hotly ever since. The Imitation Game has come under criticism within the Computer Science and Artificial Intelligence communities with leading scientists proposing alternatives, revisions, or even that the Game be abandoned entirely. Yet Turing’s imagined conversational fragments between human and machine are rich with complex instances (...) of inference of implied information, reasoning from generalizations, and meta-reasoning, challenges AI practitioners have wrestled with since at least 1980 and continue to study. We argue that the very fact the Imitation Game is so difficult may be the very reason it shouldn’t be changed or abandoned. The semi-decidability of the game at this point hints at the possibility of a hard limit to the powers of technology. (shrink)

In discussion on consciousness and the hard problem, there is an unquestioned background assumption, namely, our experience is stored in the brain. Yet Bergson argued that this very question, “Is experience stored in the brain?” is the critical issue in the problem of consciousness. His examination of then-current memory research led him, save for motor or procedural memory, to a “no” answer. Others, for example Sheldrake, have continued this negative assessment of the research findings. So, has this assumption actually been (...) proven since Bergson? Do we know how experience is stored? Or that it is stored? Here, a recent review and model of memory is examined to see where this assumption actually stands. Again, the assessment will be that nothing has changed. The core of the problem, it will be argued, lies in two things: Firstly, the search for how/where experience is stored is motivated—rephrasing Bergson—in the classic metaphysic, a framework on space and time whose logic cannot be coherently, logically adhered to in attempting to explain how experience is stored. Secondly, the search generally assumes an inadequate theory of perception that is implicitly based in this classic metaphysic. If framed within Bergson’s model of perception and his temporal metaphysic, conjoined with J. J. Gibson’s model, the storage-search appears misguided from the start. (shrink)

This chapter presents a new semantics for inductive empirical knowledge. The epistemic agent is represented concretely as a learner who processes new inputs through time and who forms new beliefs from those inputs by means of a concrete, computable learning program. The agent’s belief state is represented hyper-intensionally as a set of time-indexed sentences. Knowledge is interpreted as avoidance of error in the limit and as having converged to true belief from the present time onward. Familiar topics are re-examined within (...) the semantics, such as inductive skepticism, the logic of discovery, Duhem’s problem, the articulation of theories by auxiliary hypotheses, the role of serendipity in scientific knowledge, Fitch’s paradox, deductive closure of knowability, whether one can know inductively that one knows inductively, whether one can know inductively that one does not know inductively, and whether expert instruction can spread common inductive knowledge—as opposed to mere, true belief—through a community of gullible pupils. (shrink)

In this dissertation, I argue that understanding possesses unique epistemic value. I propose and defend a novel account of understanding that I call the management account of understanding, which is the view that an agent A understands a subject matter S just in case A has the ability to extract the relevant information and exploit it with the relevant cognitive capacities to answer questions in S. Since inquiry is the process of raising and answering questions, I argue that without understanding, (...) it would be impossible to engage in successful inquiry. I argue that understanding is indispensable for effective cognition and that it is irreducible to other epistemic categories, such as knowledge, justification, or rationality. Understanding is an irreducible component of epistemic excellence. In arguing for this account, I focus on the nature and requirements of cognition and pay less attention to intuitions about hypothetical cases, in contrast to previous approaches to understanding. I draw upon psychology and cognitive science, especially work on the frame problem, to demonstrate the importance (and difficulty) of finding relevant information and exploiting it in relevant ways to answer questions. One’s cognitive system must be organized in such a way that one has the ability to think the relevant thoughts in the relevant ways to answer the relevant questions. This reveals well-defined directions in which future research on understanding should move. In addition to proposing and defending the management account of understanding, I explore several features of understanding, including some interesting consequences of the management account. I argue that the management account and commonly used methods of measuring degrees of understanding imply that understanding can be based on false beliefs. I argue that the management account (and any account of understanding that implies understanding is an ability) implies that understanding can be acquired by luck. I also argue that understanding is ontologically prior to explanation: a good explanation is essentially an act that produces understanding in the relevant cognitive background. (shrink)

It has been just over 100 years since the birth of Alan Turing and more than 65 years since he published in Mind his seminal paper, Computing Machinery and Intelligence. In the Mind paper, Turing asked a number of questions, including whether computers could ever be said to have the power of “thinking”. Turing also set up a number of criteria—including his imitation game—under which a human could judge whether a computer could be said to be “intelligent”. Turing’s paper, as (...) well as his important mathematical and computational insights of the 1930s and 1940s led to his popular acclaim as the “Father of Artificial Intelligence”. In the years since his paper was published, however, no computational system has fully satisfied Turing’s challenge. In this paper we focus on a different question, ignored in, but inspired by Turing’s work: How might the Artificial Intelligence practitioner implement “intelligence” on a computational device? Over the past 60 years, although the AI community has not produced a general-purpose computational intelligence, it has constructed a large number of important artifacts, as well as taken several philosophical stances able to shed light on the nature and implementation of intelligence. This paper contends that the construction of any human artifact includes an implicit epistemic stance. In AI this stance is found in commitments to particular knowledge representations and search strategies that lead to a product’s successes as well as its limitations. Finally, we suggest that computational and human intelligence are two different natural kinds, in the philosophical sense, and elaborate on this point in the conclusion. (shrink)

In this paper we consider persuasion in the context of practical reasoning, and discuss the problems associated with construing reasoning about actions in a manner similar to reasoning about beliefs. We propose a perspective on practical reasoning as presumptive justification of a course of action, along with critical questions of this justification, building on the account of Walton. From this perspective, we articulate an interaction protocol, which we call PARMA, for dialogues over proposed actions based on this theory. We outline (...) an axiomatic semantics for the PARMA Protocol, and discuss two implementations which use this protocol to mediate a discussion between humans. We then show how our proposal can be made computational within the framework of agents based on the Belief-Desire-Intention model, and illustrate this proposal with an example debate within a multi agent system. (shrink)

There has been considerable work on practical reasoning in artificial intelligence and also in philosophy. Typically, such reasoning includes premises regarding means–end relations. A clear semantics for such relations is needed in order to evaluate proposed syllogisms. In this paper, we provide a formal semantics for means–end relations, in particular for necessary and sufficient means–end relations. Our semantics includes a non-monotonic conditional operator, so that related practical reasoning is naturally defeasible. This work is primarily an exercise in conceptual analysis, aimed (...) at clarifying and eventually evaluating existing theories of practical reasoning (pending a similar analysis regarding desires, intentions and other relevant concepts). (shrink)

While logical theories of information attitudes, such as knowledge, certainty and belief, have flourished in the past two decades, formalization of other facets of rational behavior have lagged behind significantly. One intriguing line of research concerns the concept of intention. I will discuss one approach to tackling the notion within a logical framework, based on a database perspective.

As many philosophers agree, the frame problem is concerned with how an agent may efficiently filter out irrelevant information in the process of problem-solving. Hence, how to solve this problem hinges on how to properly handle semantic relevance in cognitive modeling, which is an area of cognitive science that deals with simulating human's cognitive processes in a computerized model. By "semantic relevance", we mean certain inferential relations among acquired beliefs which may facilitate information retrieval and practical reasoning under certain epistemic (...) constraints, e. g., the insufficiency of knowledge, the limitation of time budget, etc. However, traditional approaches to relevance—as for example, relevance logic, the Bayesian approach, as well as Description Logic—have failed to do justice to the foregoing constraints, and in this sense, they are not proper tools for solving the frame problem/relevance problem. As we will argue in this paper, Non-Axiomatic Reasoning System (NARS) can handle the frame problem in a more proper manner, because the resulting solution seriously takes epistemic constraints on cognition as a fundamental theoretical principle. (shrink)

cuted actions. It has been applied to several challenge problems in the theory of commonsense knowledge. We study the relationship between this formalism and other work on nonmonotonic reasoning and knowl-.

The frame problem is a problem in artificial intelligence that a number of philosophers have claimed has philosophical relevance. The structure of this paper is as follows: (1) An account of the frame problem is given; (2) The frame problem is distinguished from related problems; (3) The main strategies for dealing with the frame problem are outlined; (4) A difference between commonsense reasoning and prediction using a scientific theory is argued for; (5) Some implications for the..

A plan is made to fly from Glasgow to Moscow and is shown by circumscription to lead to the traveller arriving in Moscow. Then a fact about an unexpected obstacle---the traveller losing his ticket---is added without changing any of the previous facts, and the original plan can no longer be shown to work if it must take into account the new fact. However, an altered plan that includes buying a replacement ticket can now be shown to work. The formalism used (...) is a modification of one developed by Vladimir Lifschitz, and I have been informed that the modification isn't correct, and I should go back to Lifschitz's original formalism. April 14,2001: I still haven't done it, so this article has to be regarded as tentative. I hope to fix the problems without going back to Lifschitz's formalism, which I find awkward.}. (shrink)

The Yale Shooting Problem introduced by Steve Hanks & Drew McDermott (1987) is a well-known test case of non-monotonic temporal reasoning. There is a sequence of situations. In the initial situation a gun is not loaded and the target is alive. In the next situation the gun is loaded. Eventually, a shot is fired, perhaps with fatal consequences. In this scenario there are two "fluents", alive and loaded, and two actions, load and shoot. Being loaded and being alive are inert (...) propositions in the sense that if they are true at a given moment, they will be true at the next moment unless some action such as.. (shrink)

In artificial intelligence (AI), a number of criticisms were raised against the use of probability for dealing with uncertainty. All these criticisms, except what in this article we call the non-adequacy claim, have been eventually confuted. The non-adequacy claim is an exception because, unlike the other criticisms, it is exquisitely philosophical and, possibly for this reason, it was not discussed in the technical literature. A lack of clarity and understanding of this claim had a major impact on AI. Indeed, mostly (...) leaning on this claim, some scientists developed an alternative research direction and, as a result, the AI community split in two schools: a probabilistic and an alternative one. In this article, we argue that the non-adequacy claim has a strongly metaphysical character and, as such, should not be accepted as a conclusive argument against the adequacy of probability. (shrink)

This paper presents an approach to artificial intelligence planning based on linear temporal logic (LTL). A simple and easy-to-use planning language is described, Planning Domain Description Language with control Knowledge (PDDL-K), which allows one to specify a planning problem together with heuristic information that can be of help for both pruning the search space and finding better quality plans. The semantics of the language is given in terms of a translation into a set of LTL formulae. Planning is then reduced (...) to “executing” the LTL encoding, i.e. to model search in LTL. The feasibility of the approach has been successfully tested by means of the system Pdk, an implementation of the proposed method. (shrink)