A number of examples are given of how localist models may incorporate distributed representations, without the types of nonlocal interactions that often render distributed models implausible. The need to analyze the information that is encoded by these representations is also emphasized as a metatheoretical constraint on model plausibility.

The target articles in this volume address the three major questions about dreaming that have been most responsible for the delay in progress in this field over the past 25 years. These are: (1) Where in the brain is dreaming produced, given that dream reports can be elicited from sleep stages other than REM? (2) Do dream plots have any intrinsic meaning? (3) Does dreaming serve some specialized function? The answers offered here when added together support a new model of dreaming that is testable, and should revitalize this area of study.

[Hobson et al.; Nielsen; Revonsuo; Solms; Vertes & Eastman]

Stanovich & West (S&W) appear to overlook the adaptivity of variation. Behavioral variability, both between and within individuals, is an absolute necessity for phylogenetic and ontological adaptation. As with all heritable characteristics, inter-individual behavioral variation is the foundation for natural selection. Similarly, intra-individual variation allows a broad exploration of potential solutions. Variation increases the likelihood that more optimal behaviors are available for selection. Four examples of the adaptivity of variation are discussed: (a) Genetic variation as it pertains to behavior and natural selection; (b) behavioral and cognitive aspects of mate selection which may facilitate genetic diversity; (c) variation as a strategy for optimizing learning through greater exploration; and (d) behavioral variation coupled with communication as a means to propagate individually discovered behavioral success.

We applaud Norris et al.'s critical review of the literature on lexical effects in phoneme decision making, and we sympathize with their attempt to reconcile autonomous models of word recognition with current research. However, we suggest that adaptive resonance theory (ART) may provide a coherent account of the data while preserving limited inhibitory feedback among certain lexical and sublexical representations.

Norris et al. argue against using evidence from phonetic decision making to support top-down feedback in lexical access on the grounds that phonetic decision relies on processes outside the normal access sequence. This leaves open the possibility that bottom-up connectionist models, with some contextual constraints built into the access process, are still preferred models of spoken-word recognition.

Rolls's preliminary definitions of emotion and speculative restriction of consciousness, including emotional sentience, to humans, display behaviorist prejudice. Reinforcement and causation are not by themselves sufficient conceptual resources to define either emotion or the directedness of thought and motivated action. For any adequate definition of emotion or delimitation of consciousness, new physiology, such as Rolls is contributing to, and also the resources of other fields, will be required.

An extensive survey by Heathcote et al. (in press) found that the Law of Practice is closer to an exponential than a power form. We show that this result is hard to obtain for models using leaky competitive units when practice affects only the input, but that it can be accommodated when practice affects shunting self-excitation.

This commentary brings further support for the Trace Deletion Hypothesis (TDH) from a new study of OVS (Object-Verb-Subject) and OSV (Object-Subject-Verb) sentences in Hebrew, which are active constructions that involve object movement but no change in morphology. The comprehension of these constructions in Broca's aphasia is impaired, and the performance is at chance level, as predicted by the TDH.

It has been widely demonstrated that dreaming occurs throughout human sleep. However, we once again are facing new variants of the equation “REM sleep = Dreaming.” Nielsen proposes a model that assumes covert REM processes in NREM sleep. I argue against this possibility, because dream research has shown that REM sleep is not a necessary condition for dreaming to occur.

[Nielson]

Can the understanding/acceptance principle help us to decide between alternative normative theories? There is little evidence that this principle can successfully be applied; there are no strong normative statements in Stanovich & West's target article. There is also no evidence for success of rational norms when applied to real life decisions.

Relatively poor memory for dreams is important evidence for Hobson et al.'s model of conscious states. We describe the time-gap experience as evidence that everyday memory for waking states may not be as good as they assume. As well as being surprisingly sparse, everyday memories may themselves be systematically distorted in the same manner that Revonsuo attributes uniquely to dreams.

[Hobson et al.; Revonsuo]

Our long-standing hypothesis has been that feedforward information flow is sufficient for speech perception, reading, and sentence (syntactic and semantic) processing more generally. We are encouraged by the target article's argument for the same hypothesis, but caution that more precise quantitative predictions will be necessary to advance the field.

Page argues convincingly for several important properties of localist representations in connectionist models of cognition. I argue that another important property of localist representations is that they serve as the starting point for connectionist representations of symbolic (relational) structures because they express meaningful properties independent of one another and their relations.

Rolls's presentation of emotion as integral to cognition is a welcome counter to a long tradition of treating them as antagonists. His eduction of experimental evidence in support of this view is impressive. However, we find his excursion into the philosophy of consciousness less successful. Rolls gives syntactical manipulation the central role in consciousness (in stark contrast to Searle, for whom “mere” syntax inevitably falls short of consciousness), and leaves us wondering about the roles left for emotion after all.

Both linguistic and empirical evidence fail to support Grodzinsky's account of Broca's aphasics' comprehension problems. We address concerns regarding Grodzinsky's referring to the internal subject hypothesis, the importance of case information in thematic role assignment, the processing of passives, and the adequacy of Grodzinsky's linear strategy.

The notion of “cognitive ability” leads to paradoxical conclusions when invoked to explain Inhelder and Piaget's research on class inclusion reasoning and research on the inclusion rule in the heuristics-and-biases program. The vague distinction between associative and rule-based reasoning overlooks the human capacity for semantic and pragmatic inferences, and consequently, makes intelligent inferences look like reasoning errors.

Rolls uses evolutionary theory and behavioral learning theory in his analysis of emotion. We believe that both theories are greatly underutilized, leaving an incomplete description of the nature of emotion and its neural foundation.

Stanovich & West claim that the positive correlation between reasoning tasks negates the view that errors in reasoning are due to failures in information processing. This is not correct. They conjecture that errors in reasoning are associated with conflicts between intentional and associative reasoning. This interesting proposition suggests studies relating situational characteristics to the quality of human reasoning.

Page's target article makes a good case for the strength of localist models. This can be characterized as an issue of where new information is integrated with respect to existing knowledge structures. We extend the analysis by discussing the dimension of when this integration takes place, the implications, and how they guide us in the creation of cognitive models.

Grodzinsky's localization claim can be questioned on empirical grounds. The Trace Deletion Hypothesis fails to account for a number of comprehension facts in Broca's aphasia and conduction aphasics show similar comprehension patterns. Frontoparietal systems are recruited during sentence comprehension only under conditions of increased processing load and/or attentional demands.