If reports given on experimental awakenings validly represent mental activity that was underway before the awakening, REM sleep is neither necessary nor sufficient for dreaming. Another intuitively attractive hypothesis for its function – that REM consolidates or otherwise modifies memory traces acquired while awake – is not supported by the preponderant evidence. There is growing acceptance of the possibility that REM functions to support sleep rather than waking brain processes.

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

Stanovich & West's assumption of discrete System 1 and System 2 mechanisms is questionable. System 2 can be understood as emerging from individuals who score high on several normally distributed cognitive mechanisms supporting System 1. Cognitions ascribed to System 1 and System 2 appear to be directed toward the same evolutionary significant goals, and thus are likely to have emerged from the same selection pressures.

Localist networks are symbolic models, because their nodes refer to extra-mental objects and events. Hence, localist networks can be combined with symbolic computations to form hybrid models. Such models are already familiar and they are likely to represent the dominant type of cognitive model in the next few decades.

Rolls's proposal that the amygdala is critical for the association of visual objects with reward is not consistent with recent ablation evidence. Stimulus-reward association learning is more likely to depend on basal forebrain efferents to the inferior temporal cortex, some of which pass through the amygdala. It is more likely that the amygdala is involved in rapid modulation of stimulus reward value.

Unfortunately, some researchers think a good scientific theory is one that has been repeatedly confirmed, and a bad theory is one that has not received consistent confirmation. However, confirmation of a theory depends on the extent to which a hypothesis exposes itself to disconfirmation. One confirmation of a highly specific, falsifiable experiment can have a far greater impact than the disconfirmation of twenty experiments that are virtually unfalsifiable. This commentary (1) counteracts misleading biases regarding the REM sleep/memory consolidation theory, and (2) demonstrates how chaotic cerebral activation during sleep is an essential component of long-term memory storage processes.

[Vertes & Eastman]

Norris, McQueen & Cutler claim that Merge is an autonomous model, superior to the interactive TRACE model and the autonomous Race model. Merge is actually an interactive model, despite claims to the contrary. The presentation of the literature seriously distorts many findings, in order to advocate autonomy. It is Merge's interactivity that allows it to simulate findings in the literature.

The rationality debate centers on the meaning of deviations of decision makers' responses from the predictions/prescriptions of normative models. But for the debate to have significance, the meaning and functions of normative analysis must be clear. Presently, they are not, and the debate's persistence owes much to conceptual blur. An attempt is made here to clarify the concept of normative analysis.

Rolls attributes to consciousness the functions of reflection, planning, and error-correction. Neuropsychologically grounded cybernetic theory provides an analogous, broader conceptualization: consciousness constructs goals (and plans), alters the valence of goal-related phenomena, registers error-signals, and explores unexpected circumstances (reconfiguring goals and plans as necessary). Consciousness plays a fundamental unrecognized ontological role, as well, conferring the status of “discriminable object” on select aspects of otherwise indeterminate “being.”

Grodzinsky has argued that the traces deleted in Broca's aphasia are “phonetically silent but syntactically active” (sect. 2.). If we assume such traces to be visuospatial in nature, and adopt the term “overwriting” from the author's partial matching theory (1998), we can account for the errors made by Broca's aphasics in comprehending Grodzinsky's Examples (5a), (5b), and (6).

The distinction made by Page between localist and distributed representations seems confounded by the distinction between competitive and associative learning. His manifesto can also be read as a plea for competitive learning. The power of competitive models can even be extended further, by simulating similarity effects in forced-choice perceptual identification (Ratcliff & McKoon 1997) that have defied explanation by most memory models.

Hesitations about accepting whole-heartedly Norris et al.'s suggestion to abandon feedback in speech processing models concern (1) whether accounting for all available data justifies additional layers of complexity in the model and (2) whether characterizing Merge as non- interactive is valid. Spoken word recognition studies support the nature of Merge's lexical level and suggest that phonemes should comprise the prelexical level.

Rolls demonstrates how reward/punishment systems are key mediators of cognitive appraisal, and this suggests a fundamental, causal role for emotion in thought and behaviour. However, this causal role for emotion seems to drop out of Rolls's model of consciousness, to be replaced by the old idea that emotion is essentially epiphenomenal. We suggest a modification to Rolls's model in which cognition and emotion activate each other reciprocally, both in appraisal and consciousness, thus allowing emotion to maintain its causal status where it matters most.

This commentary focuses on the implications of practical reasoning research for the view of rationality in Stanovich & West's target article. Practical reasoning does not correlate with intelligence or other reasoning tasks. Explanation in decision making terms raises the issue of dilemmas, making it hard to specify the correct norm, when an action can satisfy or conflict with two equally justifiable goals.

I wholeheartedly endorse Grodzinsky's program of attempting to tie the particular deficits observed in Broca's aphasics' comprehension and production to changes in their mentally represented model of grammar. At the level of detail, however, I see problems with two specific changes that Grodzinsky posits. One is a default Agent-assignment strategy in comprehension. The other is the hypothesis that production involves pruning all functional projections above Agreement Phrase.

The five papers in this issue all deal with the proper evolutionary function of sleep and dreams, these being different. To establish that some trait of character is an adaptation in the strict biological sense requires a story about the fitness enhancing function it served when it evolved and possibly a story of how the maintenance of this function is fitness enhancing now. My aim is to evaluate the proposals put forward in these papers. My conclusion is that although sleep is almost certainly an adaptation, dreaming is not.

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

A dynamic threshold, which controls the nature and course of learning, is a pivotal concept in Page's general localist framework. This commentary addresses various issues surrounding biologically plausible implementations for such thresholds. Relevant previous research is noted and the particular difficulties relating to the creation of so-called instance representations are highlighted. It is stressed that these issues also apply to distributed models.

While concurring that the evidence from individual differences supports a Meliorist view in the rationality debate, this commentary suggests that (1) the evidence does not clearly support a two-systems interpretation against an interpretation in terms of idiosyncratic differences in mental models, and that (2), especially where interactional processing is concerned, evidence from experimental game theory should also be considered.

Both Solms and Nielsen acknowledge the difficulty of accounting for the similarities between REM and NREM sleep mentation with a two-generator model, and each link dreams, either explicitly (Solms) or implicitly (Nielsen), to brain activation. At present, however, no data indicate that brain activation can be demonstrated whenever vivid dream reports are obtained.

[Nielsen; Solms]

The cerebral distinctness of the linguistic and mathematical faculties does not entail their functional independence. Approaches to language that posit a common foundation for the two make claims about design features, not location, and are thus not affected by the finding that one ability can be spared by a neurological accident that compromises the other.

The “reward” systems described by Rolls are systems for drive-reinforced associations of contact and distant stimuli and not for emotional behaviours. The neural systems delineated may be associated with distinct categories of “affect,” namely “hedonic feelings,” “moods,” and “emotions.” Awareness of these affects requires external perceptual as well as internal feedback. Levels of feedback in evolution and development suggest sensory qualia may not require language.