The attached report is a fascinating psychological analysis of bilingualism vs monolingualism, and effects on the brain likely to reflect epigenetic changes. According to some estimates, more than half the world’s population is multilingual –– to some extent. Because of the centrality of language use in human experience, and deep connections between linguistic and nonlinguistic processing, it would not be surprising to find that there are interactions between bilingualism and cognitive and brain processes. The difficulty is teasing out all the “background noise” in order to show (statistically significantly) this effect.
This author uses the framework of experience-dependent plasticity to evaluate evidence for systematic modifications of brain and cognitive systems that can be attributed to bilingualism. Studies –– investigating the relationship between bilingualism and cognition in infants and children, younger and older adults, and patients –– have been carried out, using both behavioral and neuroimaging methods. Although most of the research discussed in the review report some relation between bilingualism and cognitive or brain outcomes, several areas of research, notably behavioral studies with young adults, largely fail to show these effects; these discrepancies are discussed and considered in terms of methodological and conceptual issues.
A transformational change in cognitive neuroscience in recent decades has come from widespread evidence for lifelong experience-related neuroplasticity and its role in understanding brain and cognitive systems. Although it was known for a long time that enriching experiences have positive effects on rat behavior and learning, extrapolation and application of this capacity to humans, particularly adults, was not recognized until recently. Research with animals has documented the details of these adaptive brain changes and the experiences that lead to them. For example, rats reared in stimulating environments have greater cortical density and perform better in learning tasks than rats raised in standard lab cages. Similarly, rats reared in social groups show more hippocampal neurogenesis and better learning skills than rats reared in isolation.
More recently, interactions between the genome and the environment have been shown to lead to alterations in DNA methylation that allow these changes to be transmitted to the offspring of rats that had been raised in stimulating environments –– supporting the crucial role of epigenetic factors in brain and cognitive outcomes. It is abundantly clear that the environment plays a crucial role in the brain and mental development of animals.
Studies that have “failed to find differences between monolingual and bilingual groups,” in actuality have instead “found no differences”, i.e. null results, but these are often interpreted as negative results. However, “absence of evidence is not evidence of absence,” and the nature of hypothesis-testing is that “not every study will produce the same result.” The interpretation of this variability is at the very foundation of inferential statistics, which is the primary method for research in the social sciences. Research results need to be evaluated in terms of the quality of the study, and not all studies are equally sound. Clearly, the phenotype of “bilingualism” vs “monolingualism” is fuzzy, at best, and more likely represents a gradient –– as is seen in almost all gene-environment interactions. The two main issues in research on bilingualism are: validity of the designation for individuals in language groups, and validity of the task as a measure of a cognitive outcome likely to be impacted by bilingualism.
Psychol Bull March 2o17; 143: 233–262
The fascinating article by Ellen Bialystok [attached] refers to Mychasluk et al, 2012 and states “More dramatically, interactions between the genome and the environment lead to changes in DNA methylation that allow these changes to be transmitted to the offspring of rats who have been raised in stimulating environments, supporting the crucial role of epigenetic factors in brain and cognitive outcomes.”
Because, at least to me, this is such a provocative finding, I reviewed the Mychasluk et al, 2012 paper (Behavioural Brain Research, 228, 294–298). The work is substantially flawed, and does not support the authors’ conclusions nor the above statement by Ellen Bialystok: the paper does not provide evidence to support the hypothesis that epigenetic changes can be transmitted to offspring in rats.
The problems of Mychasluk et al include:
1) Only four male and four female rats were studied –– the tyranny of small numbers.
2) The pups from stimulated-male rats showed significantly less brain weight (and for which the authors provide a convoluted, non-convincing explanation).
3) The “incorrect” controls were used. Authors correctly note that “experiences that change the methylation patterns in sperm before fertilization have the potential to alter epigenetic programming of future offspring.” Given that, the appropriate ‘control’ for “a father in a stimulating environment”, is “a mother in a stimulating environment”, where there is no such potential for altering epigenetic programing in future offspring. IF genuine alteration of epigenetic programing of offspring of stimulated-female rats was observed, that should cause a re-evaluation of the current understanding of oocyte development.
4) There were two functional tests: in one (exploratory behavior), the “results” were most convincing for pups of stimulated-females versus stimulated-males; In the other test of function (geotaxis task), ability was inexplicably decreased in pups of stimulated-females (no change in pups of stimulated-males) –– another convoluted, non-convincing explanation is provided.
5) The assessment of epigenetic change was an Hpa II versus Msp I digest of brain genomic DNA, followed by 3H-dCTP extension. No actual data are shown. There is no information on the number of pups analyzed. There was no difference between pups from stimulated-male versus stimulated-female parents.
6) Although conclusions were drawn, implying changes in sperm methylation status, no studies of sperm were shown.
7) Several key experiments were not performed blinded.
I would conclude that the changes reported simply reflect the normal variation that is seen in experiments that involve small sample sizes. Perhaps there is a more compelling data set?