5 Human Cognition: Are We Really Blank Slates?

Heritable Learning Biases

Cognitive evolutionary psychology without representational nativism

Article Link: http://web.b.ebscohost.com.ezproxy.kpu.ca:2080/ehost/pdfviewer/pdfviewer?vid=5&sid=f629252b-4ea1-415e-8bac-3382c821191b%40pdc-v-sessmgr06

This article composed by Denise Cummins, Robert Cummins and Pierre Poirier emphasizes a Learning-bias-and canalization framework (LBC) to provide insight to evolutionary cognitive psychology. The proposed framework involves heritable learning biases in relationship to innate cognitive modules. To begin with, the dominant paradigm of evolutionary cognitive psychology containing massive innate cognitive modules is discussed. Followed by the introduction to the two forms of heritable learning biases. Next, criticisms of massive innate cognitive modules are presented to understand the Learning-bias-and canalization framework. Lastly, the Learning-bias-canalization framework is discussed in depth.

Introduction
Applicable evolutionary cognitive psychology requires that specific cognitive abilities must be 1) heritable and 2) quasi-independent from other heritable traits. Heritability is the capability of a trait being inherited. Whereas, the term quasi-independent refers to a trait that is not dependent upon other factors for existence. The reasoning as to why cognitive modules must be heritable is that there cannot be selection for traits that are not heritable. Traits must also be quasi-independent from other heritable traits, which suggests there must be additive genetic variance. Genetic variance refers to inheriting a particular allele from a parent which causes the phenotype (known as observable traits) to differ from the mean phenotype.

Dominant paradigm in Evolutionary cognitive psychology
The Dominant paradigm in evolutionary cognitive psychology proposes that the mind can be represented as a collection of innate cognitive modules. Tooby and Cosmides (1995) characterized innate cognitive modules as the following:
Hundreds and thousands of functional computers, typically called modules as designed to solve adaptive problems specific to our hunter-gatherer ancestors
Each module contains its own agenda
There are specialized systems for grammar, facial and emotion recognition, and locating one’s own position by estimating their direction and distance travelled rather than using landmarks or electronic navigation

The two requirements regarding evolutionary cognitive psychology can be satisfied with innate cognitive modules, as the dominant paradigm in evolutionary cognitive psychology suggests. Since innate capacities are heritable, the requirements of heritability and quasi-independence can be satisfied with innate cognitive modules. In the dominant paradigm in evolutionary cognitive psychology, the term innate does not mean to suggest that a cognitive capacity is present at birth, but rather it is encoded in the genome. In other words, the capacities of an individual is converted into a code in the genome, which pertains to a full set of chromosomes that contain all inheritable traits of an organism.
However, advocates and critics of this paradigm view it to be the main theory defining the field, which may be problematic because innate cognitive modules are heavily criticised in regards to theoretical incoherence and known facts of neurobiology.

Heritable Learning Biases
Heritability and quasi independence may be satisfied with heritable learning biases in the form of architectural or chronotropic constraints. Both of which increase canalization of specific cognitive capacities in the ancestral environment.
Chronotropic constraints refer to the neural limitations on the developmental timing of events.
Architectural constraints are physical limitations of an organism that affects the way it can process information.
Canalization is part of evolution and refers to a developmental pathway that allows an organism to develop in different situations. Chronotropic and architectural constraints increase canalization of specific cognitive abilities in the ancestral environment. As an organism develops, cognitive abilities that are highly canalized due to heritable learning biases, may result in an organism behaviorally similar to another because of various environmental triggers. To exemplify this notion, babies are able to form attachments despite the environment they are in.

Criticism of innate cognitive modules
There are two influential arguments against direct genetic encoding of knowledge, known as representational nativism.
The concept of the poverty of genetic resources proposes that an innate module will require specific synaptic connections in the cortex. However, the human genome does not have the resources to directly specify a massive amount of cortical connectivity. Human genotypes contain fewer genes than once thought (about 30,000-40,000 rather than 100,000) and among these about 20-30% may be involved in brain development. However, very few genes are involved in cortical development as they are implicated in the visual domain. Therefore if there are cognitive modules, they would be found in the cortex rather than other brain regions such as Pons.
In regards to the plasticity of the cortex, neuroscientific data suggests that the cortex is very immature at birth. The idea also suggests that if the pattern of connectivity in an area of the cortex was prespecified genetically, then it should not be possible to transplant it to another area without consequences. However, research indicates that visual cortex cells transplanted to the auditory cortex connect themselves in relevance to auditory processing.

Learning-bias-and-canalization Framework (LBC)
The learning-bias-and-canalization framework does not require a genetically representational nativism, which refers to the innate structuring of mental representations that compose knowledge. However, the framework employs minimal representational nativism in organisms. Minimal representational nativism is the idea that systems are born with a minimal set of representations necessary to learn the full set of representations, which determines a behavior. Also, the framework still satisfies the two fundamental criteria for viable evolutionary cognitive psychology.
In terms of heritability, the LBC framework assumes that there is variability in learning biases in ancestral populations. As in, when a trait proves to be adaptive, there will be a tendency for selection to disperse the bias(es) responsible for the increased canalization through the population. As an outcome, there may be a population with a highly canalized cognitive trait.
Regarding quasi independence, the framework supports the idea that networks do not necessarily need to be designed to carry out particular tasks, rather the task will select the network that has the appropriate innate properties.
The LBC is composed of a number of special features that require emphasis. The framework states that learning has its costs. The costs may include time , energy and cognitive resources. Also, learning entails mistakes and if it is done in the infantile period, care must be provided by guardians. The framework also states that learning has its advantages by allowing an organism to adapt to new environments. Therefore, learning can activate a capacity that is more canalized than one that is innate in environments where the conditions fail to trigger the development of the innate capacity. The LBC framework also allows the possibility of overcoming neural deficits due to developmental problems or trauma. The LBC also appears to allow for gradual evolution of cognition since relevant architectural and chronotropic constraints can be introduced and gradually modified to target a specific cognitive capacity. Lastly, the LBC framework is compatible with innate modules.

Testing the LBC Framework
Evolutionary cognitive psychology is problematic because brains do not leave historical records and also traits of interest have been presumably fixated. As a result, capacities can be studied in artificial creatures using simulation methods.

Interactions between evolution, development and learning discussed in the LBC framework suggest potential to allow a greater understanding of cognition, which, cognitive and developmental psychology, along with neuroscience cannot explain alone. The framework avoids massive innate cognitive modules and representational nativism. Instead, the framework enhances motivation to look for non-representational learning biases, which interact with development and the environment to produce early and highly canalized cognitive capacities.