In order to be a self-governing agent, a person must govern the process by means of which she acquires the intention to act as she does. But what does governing this process require? The standard compatibilist answers to this question all assume that autonomous actions differ from nonautonomous actions insofar as they are a more perfect expression of the agent’s agency. I challenge this conception of autonomous agents as super agents. The distinguishing feature of autonomous agents is, I argue, the (...) nonagential role they play in the formation of their intentions. I offer an account of the relevant role. (shrink)

This paper develops a new framework for thinking about how understanding comes in degrees and what it means for an agent—human or artificial—to “understand” the world. We argue that understanding a target involves appreciating its modal structure, and that growth in understanding proceeds both vertically (penetrating deeper into the levels of the target) and horizontally (appreciating more of each level’s elements). We identify four levels of understanding: mapping a target’s elements (Level 1), grasping their powers and capacities (Level (...) 2), identifying difference makers (Level 3), and apprehending necessitators (Level 4). We then apply this framework to the ongoing debate about whether AI systems possess world models, clarifying what is required for such models to be vehicles of understanding. While we remain neutral on whether current AI systems genuinely understand, we argue that our framework illuminates key issues and charts a path for future research. (shrink)

Debates concerning the units and levels of selection have persisted for over fifty years. One major question in this literature is whether units and levels of selection are genuine, in the sense that they are objective features of the world, or merely reflect the interests and goals of an observer. Scientists and philosophers have proposed a range of answers to this question. This Element introduces this literature and proposes a novel contribution. It defends a realist stance and offers (...) a way of delineating genuine levels of selection by invoking the notion of a functional unit. (shrink)

Scientists and philosophers frequently speak about levels of description, levels of explanation, and ontological levels. In this paper, I propose a unified framework for modelling levels. I give a general definition of a system of levels and show that it can accommodate descriptive, explanatory, and ontological notions of levels. I further illustrate the usefulness of this framework by applying it to some salient philosophical questions: (1) Is there a linear hierarchy of levels, with (...) a fundamental level at the bottom? And what does the answer to this question imply for physicalism, the thesis that everything supervenes on the physical? (2) Are there emergent properties? (3) Are higher‐level descriptions reducible to lower‐level ones? (4) Can the relationship between normative and non‐normative domains be viewed as one involving levels? Although I use the terminology of “levels”, the proposed framework can also represent “scales”, “domains”, or “subject matters”, where these are not linearly but only partially ordered by relations of supervenience or inclusion. (shrink)

I defend the concept of levels of mechanisms, and principled pluralism about levels more generally, against the criticisms of global eliminativists. I distinguish two ways of thinking about levels—one as waves of regularity and predictability, and the other as branches in a tree of composition—that often are conflated with one another. I show that the eliminativist’s criticisms apply only to the wave view and not to the tree view and argue that the resulting mechanistic analysis of (...) class='Hi'>levels is untouched by the eliminativists’ arguments. (shrink)

Distinctions in fundamentality between different levels of description are central to the viability of contemporary decoherence-based Everettian quantum mechanics (EQM). This approach to quantum theory characteristically combines a determinate fundamental reality (one universal wave function) with an indeterminate emergent reality (multiple decoherent worlds). In this chapter I explore how the Everettian appeal to fundamentality and emergence can be understood within existing metaphysical frameworks, identify grounding and concept fundamentality as promising theoretical tools, and use them to characterize a system of (...) explanatory levels (with associated laws of nature) for EQM. This Everettian level structure encompasses and extends the ‘classical’ levels structure. The ‘classical’ levels of physics, chemistry, biology, etc. are recovered, but they are emergent in character and potentially variable across Everett worlds. EQM invokes an additional fundamental level, not present in the classical levels picture, and a novel potential role for self-location in interlevel metaphysics. When given a modal realist interpretation, EQM also makes trouble for supervenience-based approaches to levels. (shrink)

Metainferentialism is the view that a logic is determined by its valid metainferences. It comes in two variants. The finitist version holds that logical consequence is determined by derivability relations between generalised sequents. This is always a Tarskian consequence relation. According to the transfinitist version, one cannot stop here. Instead, all the possible metainferential levels should be taken into consideration when specifying a logic. We defend the finitist version against tranfinitist criticism, clarifying, on the way, the central tenets of (...) finitist metainferentialism. By way of concluding we offer a novel generic interpretation of sequents which breaks with their usual construal as reified inferences. (shrink)

Is it possible to confuse one’s own memories with imaginings? And what about confusing one’s own imaginings with memories? The extensive literature in psychology on memory errors and confabulation suggests positive answers to these questions. However, things are more complicated, and the notion of confusion deserves a more detailed analysis. In this paper, we will do so and provide several scenarios showing that these two types of confusion can occur on two different levels: reflective (the level of self-ascription) and (...) phenomenological (the level of what it is like to be in a certain mental state). To strengthen our case, we will relate at least some of our hypothetical scenarios to known conditions affecting memory or imagination. The genuine possibility of these conditions opens the door to a systematic exploration of the implications of the falsity of the impossibility claims for the adequate account of the relationship between memory and imagination. (shrink)

Journal Name: Cognitive Linguistics Issue: Ahead of print.

We examine some assumptions about the nature of 'levels of reality' in the light of examples drawn from physics. Three central assumptions of the standard view of such levels (for instance, Oppenheim and Putnam 1958) are (i) that levels are populated by entities of varying complexity, (ii) that there is a unique hierarchy of levels, ranging from the very small to the very large, and (iii) that the inhabitants of adjacent levels are related by the (...) parthood relation. Using examples from physics, we argue that it is more natural to view the inhabitants of levels as the behaviors of entities, rather than entities themselves. This suggests an account of reduction between levels, according to which one behavior reduces to another if the two are related by an appropriate limit relation. By considering cases where such inter-level reduction fails, we show that the hierarchy of behaviors differs in several respects from the standard hierarchy of entities. In particular, while on the standard view, lower-level entities are 'micro' parts of higher-level entities, on our view, a system's macro-level behavior can be seen as a ('non-spatial') part of its micro-level behavior. We argue that this second hierarchy is not really in conflict with the standard view and that it better suits examples of explanation in science. (shrink)

This paper distinguishes and critiques several forms of pluralism about the levels of selection, and introduces a novel way of thinking about the biological properties and processes typically conceptualized in terms of distinct levels. In particular, "levels" should be thought of as being entwined or fused. Since the pluralism discussed is held by divergent theorists, the argument has implications for many positions in the debate over the units of selection. And since the key points on which the (...) paper turns apply beyond this specific issue, the paper may prove of general interest in thinking about the metaphysics of science. (shrink)

How should thought and consciousness be understood within a view of the world as being through-and-through physical? Many philosophers have proposed non-reductive, levels-based positions, according to which the physical domain is fundamental, while thought and consciousness are higher-level processes, dependent on and determined by physical processes. In this book, Kevin Morris's careful philosophical and historical critique shows that it is very difficult to make good metaphysical sense of this idea - notions like supervenience, physical realization, and grounding all fail (...) to articulate a viable non-reductive, levels-based physicalism. Challenging assumptions about the mind-body problem and providing new perspectives on the debate over physicalism, this accessible and comprehensive book will interest scholars working in metaphysics, philosophy of mind, and philosophy of science. (shrink)

Despite the ubiquity of “levels of organization” in the scientific literature, a nascent “levels skepticism” now claims that the concept of levels is an inherently flawed, misleading, or otherwise inadequate notion for understanding how life scientists produce knowledge about the natural world. However, levels skeptics rely on the maligned “layer-cake” account of levels stemming from Oppenheim and Putnam’s defense of the unity of science for their critical commentary. Recourse to layer-cake levels is understandable, as (...) it is arguably the default conception of levels in philosophy. However, relying on this conception of levels undermines the initial plausibility of general dismissals of the concept, because the problems skeptics identify within the “basic idea” of the concept “levels” are merely ones already widely acknowledged in this conception. I illustrate this “guilt by association” by looking at the embedded role of “levels” in articulating intertheoretical reductionism during the latter part of the 20th century. I conclude by suggesting a methodological framework focusing on local usage patterns as a promising means for future analysis of the levels concept. (shrink)

The subject of interspecies justice has been gaining increasing attention, but accounts of the respective responsibilities for realising interspecies justice have largely been lacking by comparison. That is, who holds which duties and what do these duties require in different contexts? Against the background of this broader research area, this paper makes some headway on this theoretical desideratum by elaborating on the different levels of burden sharing for distributive interspecies justice within the context of just habitat conservation. This paper (...) thereby builds on an account of interspecies justice concerning the just distribution of ecological space in order to elaborate on two principal scenarios of burden sharing which are not mutually exclusive but layered: (1) the burdens of maintaining interspecies justice once distributive justice is achieved; and (2) the burdens of lessening and ultimately eliminating interspecies injustices. Based on three different circumstances of scarcity and two sustainability requirements, two kinds of burdens and respective responsibilities are distinguished in this context: conservation and transition costs. (shrink)

The use of “levels of abstraction” in philosophical analysis (levelism) has recently come under attack. In this paper, I argue that a refined version of epistemological levelism should be retained as a fundamental method, called the method of levels of abstraction. After a brief introduction, in section “Some Definitions and Preliminary Examples” the nature and applicability of the epistemological method of levels of abstraction is clarified. In section “A Classic Application of the Method ofion”, the philosophical fruitfulness (...) of the new method is shown by using Kant’s classic discussion of the “antinomies of pure reason” as an example. In section “The Philosophy of the Method of Abstraction”, the method is further specified and supported by distinguishing it from three other forms of “levelism”: (i) levels of organisation; (ii) levels of explanation and (iii) conceptual schemes. In that context, the problems of relativism and antirealism are also briefly addressed. The conclusion discusses some of the work that lies ahead, two potential limitations of the method and some results that have already been obtained by applying the method to some long-standing philosophical problems. (shrink)

Levels of organization and their use in science have received increased philosophical attention of late, including challenges to the well-foundedness or widespread usefulness of levels concepts. One kind of response to these challenges has been to advocate a more precise and specific levels concept that is coherent and useful. Another kind of response has been to argue that the levels concept should be taken as a heuristic, to embrace its ambiguity and the possibility of exceptions as (...) acceptable consequences of its usefulness. In this chapter, I suggest that each of these strategies faces its own attendant downsides, and that pursuit of both strategies (by different thinkers) compounds the difficulties. That both kinds of approaches are advocated is, I think, illustrative of the problems plaguing the concept of levels of organization. I end by suggesting that the invocation of levels may mislead scientific and philosophical investigations more than it informs them, so our use of the levels concept should be updated accordingly. (shrink)

Cognitive science began as a multidisciplinary endeavor to understand how the mind works. Since the beginning, cognitive scientists have been asking questions about the right methodologies and levels of explanation to pursue this goal, and make cognitive science a coherent science of the mind. Key questions include: Is there a privileged level of explanation in cognitive science? How do different levels of explanation fit together, or relate to one another? How should explanations at one level inform or constrain (...) explanations at some other level? Can the different approaches to the mind, brain, and culture be unified? The aim of this issue of topiCS is to provide a platform for discussing different answers to such questions and to facilitate a better understanding between the different strands of thinking about the right levels of explanation in cognitive science. (shrink)

Moral sensitivity has generally been interpreted in a normative sense, as the ability to notice moral features present in a situation. This paper outlines an alternative, descriptive conception of moral sensitivity: the levels of moralisation model. This model describes four qualitatively distinct levels at which a preference can be held: no moralisation; moralisation for the self; moralisation for others; and public expression of moralisation. Empirical research supporting the existence of these levels as well as processes that move (...) a preference across the levels are discussed. In addition, the practical utility of the model is demonstrated with regard to moral education and conflict resolution programs. (shrink)

The paper addresses a question concerning George Ellis’s theory of top-down causation by considering a challenge to the “level-picture of nature” which he employs as a foundational element in his theory. According to the level-picture, nature is ordered by distinct and finitely many levels, each characterised by its own types of entities, relations, laws and principles of behavior, and causal relations to their respective neighbouring top- and bottom-level. The branching hierarchy that results from this picture enables Ellis to build (...) his model of modular hierarchical structure for complexity, his account of same-level, bottom-up and top-down causation, of emergence, equivalence-classes and multiple realisability. The three main arguments for the level-picture in Ellis’s works are reconstructed and shown to face serious problems. Finally, the paper presents a possible solution to this challenge by introducing a reformulation of certain fundamental points of Ellis’s theory that does without the level-picture of nature. This allows us to preserve all of his central claims about the model of complexity, the three types of causation, emergence, equivalence-classes and multiple realisability. Any problems pertaining to the level-picture can be remedied in the context of Ellis’s theory of top-down causation. (shrink)

I consider a problem for functional reductionism, based on the following tension. Say that b is functionally reduced to a. On the one hand, a and b turn out to be identical, and identity is a symmetric relation. On the other hand, functional reductionism implies that a and b are asymmetrically related: if b is functionally reduced to a, then a is not functionally reduced to b. Thus, we ask: how can a and b be asymmetrically related if they are (...) the same thing? I propose a solution to this tension, by distinguishing between ontological levels and levels of description. (shrink)

The levels of selection problem was central to Maynard Smith’s work throughout his career. This paper traces Maynard Smith’s views on the levels of selection, from his objections to group selection in the 1960s to his concern with the major evolutionary transitions in the 1990s. The relations between Maynard Smith’s position and those of Hamilton and G.C. Williams are explored, as is Maynard Smith’s dislike of the Price equation approach to multi-level selection. Maynard Smith’s account of the ‘core (...) Darwinian principles’ is discussed, as is his debate with Sober and Wilson (1998) over the status of trait-group models, and his attitude to the currently fashionable concept of pluralism about the levels of selection. (shrink)

The most widely known account of levels of explanation in neuroscience and cognitive science is due to Marr), who proposed a three-level framework for explanation of the visual system. Various philosophers have given interpretations of Marr’s framework. The purpose of this chapter is different: to show how the analogy with artifacts (i.e., systems engineered by humans) is foundational for the idea that complex neuro-cognitive systems are amenable to explanation at distinct levels; to show how this works as a (...) simplifying strategy attractive to anti-reductionists; and to argue that the resulting explanations are imbued with an ‘all-too-human’ character. Section 1 surveys various proposals for carving the brain into levels, and describes how epistemic conceptions of levels of abstraction differ from metaphysical conceptions of levels of organisation. Section 2 examines the artifact analogies. The most salient source of analogy is of course the digital computer, but the chapter discusses how more general comparisons with processes of design and engineering also play a significant role. It shows how the analogies, and the subsequent notion of a distinct computational level, have engendered common ideas about how to safely abstract away from the complexity of concrete neural systems, yielding explanations of how neural processes give rise to cognitive functions. Section 3 raises worries about these explanations, due to neglected differences between human-made devices and biological organs. The chapter concludes with some general observations about how reductionist and anti-reductionist approaches to mind and brain stem from intuitions deriving from the human experience of making functional objects. (shrink)

We investigate a theory of Frege structures extended by the Myhill-Flagg hierarchy of implications. We study its relation to a property theory with an approximation operator and we give a proof theoretical analysis of the basic system involved. MSC: 03F35, 03D60.

It is currently common to conceive of the classic methodological individualism–holism debate in level terms. Accordingly, the dispute is taken to concern the proper level of explanations in the social sciences. In this paper, I argue that the debate is not apt to be characterized in level terms. The reason is that widely adopted notions of individualist explanations do not qualify as individual-level explanations because they span multiple levels. I defend this claim relative to supervenience, emergence, and other accounts (...) of the social world as levelled. Moreover, I discuss the consequences of this finding for the ongoing methodological individualism–holism debate. (shrink)

According to the H-Evo-Devo Theory, macroevolution is decoupled from microevolution. Microevolution belongs to the subspecies level, whose character states correspond to simple characters, e.g., color, pattern, size, and proportion. In contrast, macroevolution belongs to the species level, whose character states correspond to complex characters or body plans. This chapter explores the biological processes surrounding the origin, stability, and extinction of both subspecies and species. Concerning their origin, subspecies arise through Darwinian mechanisms (dispersal models), whereas species emerge through supraDarwinian mechanisms (prospecies (...) model). Regarding stasis, the stability of both subspecies and species is attributed to developmental constraints acting on simple and complex characters, respectively. Finally, differences in extinction are explained by a hierarchical structure: Subspecies extinction results from factors altering their local habitat and their population structure, while species extinction arises from the loss of parts and mass extinction events. Essentially, this chapter marks the first modern effort to distinctly delineate evolution into at least two levels: microevolution and macroevolution. (shrink)

This paper explores the role of phenomenology in the understanding of the cognitive processes of coupling/decoupling, defending the Wittgensteinian idea that phenomenology can play a crucial role as a description of immediate (social) experience. We argue that epistemic feelings can provide a phenomenological description of the development of a subject's everyday experience, tracking the transition from the processes of coupling/decoupling and recoupling with the world. In particular, the feeling of familiarity, whose key features can be considered the core of epistemic (...) feelings, signals a novelty in the flow of experience that makes sense and is worthy of remarking on or even articulating. By describing the primary features and sources of the feeling of familiarity, we highlight a conceptual tension related to its sources, which could be based on processing both fluency and discrepancy. We proposed a solution to the conceptual tension by introducing two levels of the feeling of familiarity: epistemic and experiential. (shrink)

We give a general background describing how the notion of levels of reality comes about in contemporary nonreductive approaches to the special sciences, what the notion of levels means in these approaches, its role in our scientific outlook of the world, and why it is important.

According to the H-Evo-Devo Theory, megaevolution belongs to the higher taxa level, and modular-evolution pertains to the homologue level. This chapter explores the biological processes surrounding the origin, stability, and extinction of higher taxa as well as the origin and stability of homologues as modules of comparative anatomy. Concerning the origin of higher taxa, the taxonomic structure in which they arise is built from the top-down, i.e., higher-level body plans originate first, followed sequentially by lower-level body plans. Additionally, the sudden (...) origin of these entities at various taxonomic ranks (as hopeful monsters) is highlighted. Regarding stability, higher taxa exhibit supraspecific stasis, a phenomenon described in other studies under the term “living fossils,” which is explained by developmental constraints. With respect to higher taxa extinction, this process results from mass extinctions and the loss of parts. Meanwhile, modules are produced through the phenomenon of co-option, and the cause of their stability is attributed to developmental constraints acting at the deepest level of the hierarchy. Overall, this chapter specifically emphasizes the biological causes at higher levels of the hierarchy. (shrink)

The meanings of words are not permanent but change over time. Some changes of meaning are quick, such as when a pronoun changes its reference; some are slower, as when two speakers find out that they are using the same word in different senses; and some are very slow, such as when the meaning of a word changes over historical time. A theory of semantics should account for these different time scales. In order to describe these different types of meaning (...) changes, I present an analysis of three levels of communication: instruction, coordination of common ground and coordination of meaning. My first aim is to show that these levels must be considered when discussing lexical semantics. A second aim is to use the levels to identify the communicative roles of some of the main word classes, in particular nouns, adjectives, verbs, indexicals and quantifiers. I argue that the existence of word classes can, to a large extent, be explained by the communicative needs that arise on the different levels. (shrink)

The article is devoted to consideration of freedom concepts as they have been developed in the history of philosophy and social practices. The relation between imagination about freedom and the forms of person's self-realization, self-perfection and self-determination is argued. On the base of theoretical achievement of M. Adler School, the author ascertains communicative network between both empirical, axiological, existentional levels of freedom and economical, ethical-political and cultural forms of social activity of people. Key words: freedom, self-realization, self-perfection, self-determination, self-affirmation, (...) vocation. cultural membership, ethical. (shrink)

By linking the Lacanian symbolic to the brain, one finds that the logic of symbolic organization applies to more than just language per se. Lacan himself is clear to articulate a particular linguistic organization to the unconscious which is not equated with specific language. Hence, the unconscious is structured like a language. This chapter sketches levels of the symbolic—different domains of brain systems structured like a language—including basic emotions, the motor axis, language motor dynamics, motoric representations, and abstract cognitive (...) control. (shrink)

Chapter 3 provides an account of levels that is applicable to all mechanisms governing natural and social phenomena. Several criteria, akin to axiomatic definitions, are introduced to categorise various things in a mechanism into different levels. These definitions use the notions of componency, direct component and ‘being at the same level’, and the chapter shows that these notions are derived from scientific practice.

In Chapter 4, the notion of ‘being at the same level’ is further defined using the notions of production (thick causation) as discussed by Gillett, Glennan and other mechanists and conditional independence as introduced by James Woodward. The goal here is to come up with an objective set of criteria grounded in scientific practice based on which one could identify the levels in a mechanism and the things present at each level. This chapter also discusses in detail the interventionist (...) framework developed by Woodward and shows how interventionist (thin) causation is different from production (thick causation) in the new mechanist framework. It also discusses to which things in mechanisms interventions can be introduced, whether the variables used in interventionist causal models strictly refer to things in mechanisms and whether causation can exist between unrelated mechanisms. This chapter shows how interventionism, as a methodological thesis about causation, is useful in identifying the presence of downward causation in mechanisms. (shrink)

I analyze the main characteristics of the hierarchy of levels of selection as manifestors of adaptation (i.e., hierarchy of levels of evolutionary organization) and show that, while ranked, this hierarchy is nonnested. Secondly, I consider a potential objection to my approach derived from Eronen and Ramsey’s (2025) claim that levels of organization should be teased apart from levels of selection. I argue that their approach rests on a failure to distinguish between interactors and manifestors of adaptation, (...) for while a hierarchy of interactors does not correspond to a hierarchy of levels of organization, a hierarchy of manifestors does. (shrink)

The idea of levels of organization plays a central role in the philosophy of the life sciences. In this article, I first examine the explanatory goals that have motivated accounts of levels of organization. I then show that the most state-of-the-art and scientifically plausible account of levels of organization, the account of levels of mechanism proposed by Bechtel and Craver, is fundamentally problematic. Finally, I argue that the explanatory goals can be reached by adopting a deflationary (...) approach, where levels of organization give way to more well-defined and fundamental notions, such as scale and composition. (shrink)

This chapter surveys some issues about levels in the sciences as they surface in the social and behavioral sciences. However, it is hard to discuss issues in the social sciences without also discussing the issues about levels in science in general. Thus, the chapter proposes a general take on debates about levels that it calls naturalist and contextualist: naturalist in that conceptual metaphysics cannot trump science and contextualist in that the use and usefulness of level concepts means (...) different things with different plausibility according to scientific context. The chapter argues that the levels concept is neither incoherent or always useful, that many objections to levels such as doubts about downward causation are anti-naturalist conceptual metaphysical approaches that should be rejected, and that levels cannot be replaced with concepts of mechanisms or scale. All this is tied down and expanded with discussions of some empirical social methodological and research areas. Section 1 sets up the general naturalist and contextualist framework, Section 2 applies these to general debates about levels, and Section 3 discusses in detail some social science research. (shrink)

In The Levels of Selection (Brandon, 1984), Robert Brandon provides a suggestive but ultimately unsuccessful attempt to use the probabilistic notion ofscreening off in providing a schema for dealing with an aspect of the units of selection question in the philosophy of biology. I characterize that failure, and suggest a revision and expansion of Brandon's account which addresses its key shortcoming.

According to Critical-Level Views in population axiology, an extra life improves a population only if that life’s welfare exceeds some fixed ‘critical level.’ An extra life at the critical level leaves the new population equally good as the original. According to Critical-Range Views, an extra life improves a population only if that life’s welfare exceeds some fixed ‘critical range.’ An extra life within the critical range leaves the new population incommensurable with the original. -/- In this paper, I sharpen some (...) old objections to these views and offer some new ones. Critical-Level Views cannot avoid certain Repugnant and Sadistic Conclusions. Critical-Range Views imply that lives featuring no good or bad components whatsoever can nevertheless swallow up and neutralise goodness or badness. Both classes of view entail that certain small changes in welfare correspond to worryingly large differences in contributive value. -/- I then offer a view that retains much of the appeal of Critical-Level and Critical-Range Views while avoiding the above pitfalls. On the Imprecise Exchange Rates View, the quantity of some good required to outweigh a given unit of some bad is imprecise. This imprecision is the source of incommensurability between lives and populations. (shrink)

Attention has been given to cross-cultural differences in adolescent growth, but far less is known about developmental variability during juvenility (ages 3–10). Previous research among the Pumé, a group of South American foragers, found that girls achieve a greater proportion of their adult stature during juvenility compared with normative growth expectations. To explain rapid juvenile growth, in this paper we consider girls’ activity levels and energy expended in subsistence effort. Results show that Pumé girls spend far less time in (...) subsistence tasks in proportion to their body size compared with adults, and they have lower physical activity levels compared with many juveniles cross-culturally. Low activity levels help to explain where the extra energy comes from to support rapid growth in a challenging environment. We suggest that activity levels are important to account for the variation of resource and labor transfers in mediating energy availability. (shrink)

Recent work in the methodology of connectionist explanation has I'ocrrsccl on the notion of levels of explanation. Specific issucs in conncctionisrn hcrc intersect with rvider areas of debate in the philosophy of psychology and thc philosophy of science generally. The issues I raise in this chapter, then, are not unique to cognitive science; but they arise in new and important contexts when connectionism is taken seriously as a model of cognition. The general questions are the relation between levels (...) and the status of levels which have no obvious relation to others. In speaking of levels, what is the connection, if there is one, between explanation and ontology? Which, if any, conccpt of reduction is applicable to connectionist systems? What kind of legitinrtcy can the constructs of common sense psychology, or of that vclsion ol intentional realism represented by classical symbol-systems n I, hirvc irr ir full-scale connectionist theory of mind? (shrink)