After reviewing historical aspects of brain evolution, this accessible book provides an enjoyable overview of several general principles of brain evolution, culminating in discussions of mammalian and human brains and a framework for future research.

van der Velde's &de Kamps's neural blackboard architecture is similar to “activation trace diagrams” (Clancey 1999), which represent how categories are temporally related as neural activations in parallel-hierarchical compositions. Examination of other comprehension examples suggests that a given syntactic categorization (structure assembly) can be incorporated in different ways within an open composition by different kinds of anchoring relations (delay assemblies). Anchors are categorizations, too, so they cannot be reused until their containing construction is completed (bindings are resolved).

Striedter's accessible concept-based book is strong on the macroevolution of brains and the developmental principles that underlie how brains evolve on that scale. In the absence of greater attention to microevolution, natural selection, and sexual selection, however, it is incomplete and not fully modern on the evolution side. Greater biological integration is needed.

We argue that van der Velde's & de Kamps's model does not solve the binding problem but merely shifts the burden of constructing appropriate neural representations of sentence structure to unexplained preprocessing of the linguistic input. As a consequence, their model is not able to explain how various neural representations can be assigned to sentences that are structurally ambiguous.

The tension between focusing on species similarities versus species differences (phylogenetic versus adaptationist approaches) recurs in discussions about the nature of neural connectivity and organization following brain expansion. Whereas Striedter suggests a primary role for response inhibition, other possibilities include dense recurrent connectivity loops. Computer simulations and brain imaging technologies are crucial in better understanding actual neuronal connectivity patterns.

Increases of absolute brain size during evolution reinforced stronger structuring of brain connectivity. One consequence is the hierarchical cluster structure of neural systems that combines predominantly short, but not strictly minimal, wiring with short processing pathways. Principles of “large equals well-connected” and “minimal wiring” do not completely account for observed patterns of brain connectivity. A structural model promises better predictions.

van der Velde & de Kamps make a case for neural blackboard architectures to address four questions raised by human language. Unfortunately, they neglect a sizable literature relating blackboard architectures to other fundamental cognitive questions, specifically consciousness and voluntary control. Called “global workspace theory,” this literature integrates a large body of brain and behavioral evidence to come to converging conclusions.

The authors, on the basis of brief arguments, have dismissed tensor networks as a viable response to Jackendoff's challenges. However, there are reasons to believe that connectionist approaches descended from tensor networks are actually very well suited to answering Jackendoff's challenges. I rebut their arguments for dismissing tensor networks and briefly compare the approaches.

Brain evolution is a complex weave of species similarities and differences, bound by diverse rules and principles. This book is a detailed examination of these principles, using data from a wide array of vertebrates but minimizing technical details and terminology. It is written for advanced undergraduates, graduate students, and more senior scientists who already know something about “the brain,” but want a deeper understanding of how diverse brains evolved. The book's central theme is that evolutionary changes in absolute brain size tend to correlate with many other aspects of brain structure and function, including the proportional size of individual brain regions, their complexity, and their neuronal connections. To explain these correlations, the book delves into rules of brain development and asks how changes in brain structure impact function and behavior. Two chapters focus specifically on how mammal brains diverged from other brains and how Homo sapiens evolved a very large and “special” brain.

Evolutionary theory and methods are central to understanding the design of organisms, including their brains. This book does much to demonstrate the value of evolutionary neuroscience. Further work is needed to clarify the ways that neural systems evolved in general (specifically, the interaction between mosaic and coordinated evolution of brain components), and phylogenetic methods should be given a more prominent role in the analysis of comparative data.

Because network scaling costs tend to limit absolute brain size, Striedter suggests that large cetacean brains must have evolved some novel ways to cope with these costs. A new analysis of available data shows that the scaling pattern of interhemispheric connectivity in cetaceans is isometric and differs from that observed in terrestrial mammals.

Overall, most of the reviewers agree that Principles of Brain Evolution was a welcome addition to the field, and kindly describe it as carefully researched and lucidly written. Thereafter, they note some gaps – principally, adaptive scenarios, microevolutionary studies, and computational models. I here admit to those deficiencies but explain why they exist and how they might be filled. In addition, one commentator criticizes my analysis of hominin brain evolution, and another finds my principle of “large equals well-connected” to be inconsistent with the data. I rebut those two critiques. Hopefully, this process of critique and counterpoint will stimulate some readers to pursue the mentioned thoughts and to engage in new research.

The target article presents a stimulating account for some of the most challenging issues in cognitive neuroscience. The theory solves in neural terms cognitive problems beyond the scope of previous models. But in many aspects the neural implementation is a quite literal translation of symbolic descriptions and therefore still lacks some of the truly self-organizing properties characteristic of biological networks.

Various issues concerning the neural blackboard architectures for combinatorial structures are discussed and clarified. They range from issues related to neural dynamics, the structure of the architectures for language and vision, and alternative architectures, to linguistic issues concerning the language architecture. Particular attention is given to the nature of true combinatorial structures and the way in which information can be retrieved from them in a productive and systematic manner.

van der Velde's & de Kamps's model encodes complex word-to-word relations in sentences but does not encode the hierarchical constituent structure of sentences, a fundamental property of most accounts of sentence structure. We summarize what is at stake and suggest two ways of incorporating constituency into the model.

I discuss weaknesses of the proposed model related to reinstantiation of encodings recorded by the hippocampal complex and to the inability of the model to explain complexity phenomena. An alternative model that also addresses the formation of hierarchical representations of sentences in working memory is outlined, and the ability of this model to account for complexity phenomena is briefly reviewed.

We question the ecological plausibility as a general model of cognition of van der Velde's & de Kamps's combinatorial blackboard architecture, where knowledge-binding in space and time relies on the structural rules of language. Evidence against their view of the brain and an ecologically plausible, alternative model of cognition are brought forward.

Selection for or against muscle initiates a cascade of centripetal (outside-in), trophically mediated, neurological events through which the environment programs heritable neuromuscular and neuroneuronal connections in a rapid and specific fashion. The velocity, direction, and efficiency of this process are a consequence of the environment acting directly on muscle, the organ of action, and behavioral interface between organism and environment.

Cognitive function certainly poses the biggest challenge for computational neuroscience. As we argue, past efforts to build neural models of cognition (the target article included) had too narrow a focus on implementing rule-based language processing. The problem with these models is that they sacrifice the advantages of connectionism rather than building on them. Recent and more promising approaches for modeling cognition build on the mathematical properties of distributed neural representations. These approaches truly exploit the key advantages of connectionism, that is, the high representational power of distributed neural codes and similarity-based pattern recognition. The architectures for cognitive computing that emerge from these approaches are neural associative memories endowed with additional mapping operations to handle invariances and to form reduced representations of combinatorial structures.

Striedter's account of human brain evolution fails on two key counts. First, he confuses developmental constraints with selection explanations in the initial jump in hominid brain size around two MYA. Second, he misunderstands the Machiavellian Intelligence explanation.