The application of neurogenomics to education: analyzing guiding visions

Abstract

The field of neurogenomics, which studies the role of genomics in the development and function of the nervous system, is considered by many scholars to have the potential to contribute to learning and education. In order to gain more insights into the opportunities and concerns associated with this potential, this paper explores scientists' guiding visions for this domain. In addition to an exploration of relevant literature, one focus group and six interviews were conducted with neurogenomics scientists, with the results showing that scientists working in different areas of neurogenomics formulate diverse guiding visions and that most guiding visions are clinically oriented. We demonstrate how these guiding visions originate from scientists' disciplinary background, empirical practice, overarching theories and moral appreciations, giving rise to conflicting understandings of key concepts such as “phenotype”, “intelligence” and “environment”. This variance indicates that more interdisciplinary communication and cooperation is needed before steps are made in the direction of education.

Keywords:

• neurogenomics
• education
• vision assessment

Introduction

The interdisciplinary field of neuroscience contributes to medical practice through the prevention and diagnosis of disorders at the same time as providing increasing insights into the biological pathways of the brain. The development of neuroscience has given rise to a number of new disciplines, including the emerging field of “neuroeducation” which seeks to understand the link between neural pathways and learning (Frazzetto and Anker 2009). Some scholars consider that improved understanding of memory, attention and stress may eventually have practical applications within the classroom (for instance, Fischer 2009; Kuhl 2011; Pasquinelli 2011).

In the Netherlands, research linking education and cognition has received a great deal of media attention. One example of this is the debate surrounding differences in brain development between boys and girls and the possible consequences for educational policy (see e.g. Willems 2009). Internationally, the field has seen interesting developments over the past few years, such as the foundation of the International Mind, Brain and Education Society in 2004 and the subsequent launching of their journal Mind, Brain and Education (Carew and Magsamen 2010). These developments have led to the perception that cooperation between researchers and educators may be mutually beneficial with steps being taken to translate neuroscientific findings into improved instructional strategies and learning environments (Society for Neuroscience 2009).

In their review of the emerging field of educational neuroscience in the journal Neuroethics, Ansari, De Smedt, and Grabner (2012) provide an overview of key advances in neuroscientific studies relevant to learning. They note that the future of educational neuroscience may involve “collaborations with geneticists in an effort to understand the contribution of genetic variability to measures of cognitive processes or the relationship between genetic markers and brain activation during cognitive tasks” (3). However, the direction of this collaboration between neuroscience and genomics is still unclear. In this article, we therefore focus on the possible contribution of the field of neurogenomics to the educational sector.

A number of scholars have predicted that neurogenomics could lead to the development of applications that have implications for the education of the future. For example, neurogenomics might contribute to the development of new therapies for learning disorders (Wermter et al. 2010) and predict high learning ability (Haworth et al. 2007) or the need for the provision of tailored tuition (Schulte-Körne et al. 2007). Society also has high hopes for the contribution of neuroscience and genetics to education (Fischer et al. 2007). However, these hopes are not always realistic and neuroscientists have warned that it is necessary to manage expectations (Bruer 1997; Ansari, De Smedt, and Grabner 2012). At the same time, the field of neurogenomics is a very controversial area of neuroscience given its ethical, social and legal implications relating to discrimination, privacy and confidentiality, as demonstrated in a study of US and UK newspaper articles (Racine, van der Loos, and Illes 2007). Anticipation of the promise of these new developments and identification of pitfalls can assist in stimulating the development of desirable applications.

The case of neurogenomics is, in our opinion, a clear example of a Collingridge dilemma: technology development is hard to control in its early phases because the social consequences cannot be predicted, while, in later phases, the technology becomes entrenched in society and steering becomes very difficult (1981, 11). In the current phase of the development of neurogenomics, there are no concrete applications in the domain of education to help understand its implications. We therefore follow the example of Roelofsen et al. (2008) who have assessed the guiding visions of scientists in the field of ecogenomics as one way of trying to understand how this field is likely to develop. Guiding visions can be described as mental images of desirable and attainable futures that are shared by a collection of actors (Grin and Grunwald 2000). Expectations and visions in science and technology are important objects of study because they generate the activities that shape innovation, mobilizing resources at the levels of both policy (national and institutional) and the individual researcher (Borup et al. 2006).

As proposed by Illes and Bird (2006), we also want to gain insights into the assumptions underlying these visions. For this purpose, we use the distinction between first- and second-orders of discourse identified by Grin and Van de Graaf (1996) (based on Fischer 1980; Schön 1995). We argue that guiding visions can be viewed as first-order discourse, which is used mostly in daily practice and consists of the problem definitions and the preferred solutions shared by actors (Grin and Van de Graaf 1996). In this study, first-order discourse relates to the issues scientists wish to overcome by means of their research, the working hypotheses they formulate in the process of research and the barriers scientists expect to face with their research.

Underlying first-order discourse is second-order discourse, consisting of the underlying assumptions and values on which researchers base their guiding visions. Researchers' assumptions are part of the overarching empirical theories they adhere to, often related to their disciplinary background. For example, Torgersen (2009) demonstrates that basic assumptions concerning the understanding of key mechanisms and concepts within the multidisciplinary field of genomics can differ enormously between technical scientists and experimental biologists: the former have a more deterministic view because they aim for control based on relevant parameters. Underlying values are a part of the researchers' appreciative system and are also typically shared by professional communities (Grin and Van de Graaf 1996). Second-order notions are generally held constant as they are part of professional training (Grin and Van de Graaf 1996).

In this article, we wish to gain more insight into scientists' guiding visions of neurogenomics by exploring their underlying second-order notions. This involves considering their empirical theories and appreciative systems.

Method

This study explores a newly emerging field and we have used three different approaches in order to gain insight into the field from different angles. First, we conducted a literature review in order to investigate current thinking on the future application of genomics in the domain of education. However, given that guiding visions are often implicit in science, we also deemed it important to speak to scientists engaged in genetics research with potential relevance to education. Therefore, we conducted exploratory, in-depth interviews to investigate scientists' visions of the future. In interviews, individual researchers were hesitant to make statements about the future. For this reason, we also conducted a focus group discussion with six scientists in order to encourage mutual reflection and stimulate joint construction of guiding visions.

Our first step was to identify guiding visions in the academic literature. In new fields, the absence of well-defined keywords hampers bibliographic searches (Suraud et al. 1995). Therefore, an extensive search query (query 1) was used to identify articles relevant to neurogenomics and education, adapted from an extensive bibliographical study of the development of the fields of neuroscience and education (Merkx et al. 2009), with the addition of key words for genomics. This search was expanded by a manual snowballing search for articles published in 2009 or 2010 within the databases, Pubmed, Web of Science® and Embase (query 2).

The literature analysis was followed by a qualitative exploration of these visions with experts in order to gain insight into the role of second-order notions. First, we conducted interviews with neurogenomics researchers working in areas relevant to education. We aimed to recruit a participant sample with maximum variety (Morse 1994) and gather data that were “information rich” (Patton 1990). To help us decide on which disciplines to include in this exploratory study, we used the definition of genomics by Boguski and Jones (2004), namely as comprising research on the role of genes in the evolution, development, structure and function of the nervous system. Participating scientists held a minimum of a post-doctoral position and worked at five different universities in the Netherlands. In total, 10 researchers were asked to participate and 6 consented to be interviewed. The interviews, which took between 60 and 90 minutes, focused on participants' opinions of how neurogenomics could potentially contribute to education and the technical, organizational and social problems that might be expected.

After the interviews, we used a similar recruitment strategy to organize a focus group of researchers in order to stimulate co-construction of future visions. The focus group did not include any of the same researchers as the interviews. We invited 18 researchers, 6 of whom attended the focus group session. The focus group started with the collection and sharing of predictions about the possible implications of neurogenomics for educational practice in the year 2040. We chose the far future as the focus for discussion so that researchers could speculate about the future direction of their research without the constraint of concrete feasibility. The participants were asked to identify desirable future applications of neurogenomics in the domain of education and were encouraged to make their visions explicit, in order to understand why this future vision was considered desirable and learn more about the second-order notions underlying these visions. Next, the participants prioritized their visions and discussed why some visions took precedence over others. This design is based on Kupper et al.'s (2007) description of the “Value Lab”, a focus group design aimed to identify the values at stake for participants by collecting, clustering and prioritizing associations of participants with a certain topic. By using a similar design, we aimed to go beyond the identification of problems for which neurogenomics could provide solutions to learn about the participants' second-order worldviews and normative ideas.

Given that this is an explorative study, we used open coding to analyze the data. In addition, we used the discourse framework of Grin and Van de Graaf (1996) to structure the guiding visions of the researchers, labeling the data as first or second-order notions. All participants were asked for their permission to record the sessions which were transcribed in full. ATLAS.ti was used for the analysis of both the interviews and focus group discussion.

Results

In this section, we first describe the different guiding visions encountered in the literature where explicit reference to second-order notions is generally absent. Next, we describe the guiding visions of scientists expressed during the interviews and focus group discussion, paying attention to both the first and second-order framing of these visions.

Literature Review

Query 1 in the Web of Science® database yielded 8530 articles published between 2001 and 2010. However, only two of the articles contained the word “education” in its title. The 201 articles with the word “learning” in the title all relate to learning disabilities and no future visions beyond the clinical setting were identified within this sample. Query 2, a manual search of the Pubmed, Web of Science® and Embase databases, yielded 29 articles relevant to the fields of neurogenomics and education. Of these, 19 contained statements about the practical implications of research discoveries for the field of education. These articles were published in journals in different interdisciplinary areas, including psychiatry and child psychiatry, neuroscience, genomics and educational neuroscience. The future visions found are presented in Table 1. Some articles contained multiple visions.

Table 1. Overview of guiding visions found in 2009–2010 publications.

Possible application		No.
Improved understanding of		19
	Basic neuroscience	1
	Biological pathways	6
	Vulnerability and resilience	1
	Heterogeneity in mental disorders	1
	Dyslexia	1
	The neural substrates of attention
	“Missing heritability”	1
	The processes involved in language and learning disorders	1
	The causal links between genes and behavior	2
	The anatomical complexity of the brain	1
	Learning disabilities	1
	Communication	1
	Pathophysiology	1
	Individual differences between children	1
New definitions of disorders		3
	Identification of further syndromes	1
Diagnosis		9
	Early diagnosis	2
	Of ADHD	1
	Of dyslexia	1
	Refinement of diagnostic criteria	1
Prevention		7
	Of learning disabilities	1
	Screening for dyslexia	1
	Family planning	1
Treatment		6
	New therapeutic strategies	1
	Individualized treatment	2
	Of ADHD	1
	Prediction of response to treatment	1
	Targeting of existing therapies	2
Prediction		4
	Of educational progress	1
	Of learning problems	1
	Providing tailored tuition	2
	Prognosis of disorder	1
New definitions of disorders		3
	Identification of further syndromes	1

Most problem statements that have been identified in this literature relate to the lack of understanding of the genetic contribution to brain function. Other future visions concern the diagnosis, treatment and prevention of disorders, such as developmental dyslexia, autism and attention deficit hyperactivity disorder (ADHD). These results show that the field has a strong clinical orientation. For example, the Journal of European Child and Adolescent Psychiatry focuses on mental disorders and future visions include “screening” and “therapeutic strategies”:

Upon discovering genes for DD [developmental dyslexia] susceptibility and their underlying causal variants, it is envisaged that 1-day young people may be screened for their potential risk in developing DD. Appropriate action may then be taken to reduce this risk by providing tailored tuition governed by their underlying genetic makeup. (Scerri and Schulte-Körne 2010, 192)

Future visions concerning the implications for the practice of education are generally absent. Where these visions are present, they have been published in the journal Mind, Brain and Education with the one exception cited above which was published in another journal. Neurogenomics interventions are generally focused on children with a learning disability. For example, it may be possible to identify children with a disability before problems become severe. According to Haworth et al. (2007, 179),

Interventions will rely on environmental engineering, such as teaching and classroom interventions, not on genetic engineering, which is not possible for complex traits that involve many genes of small effect size.

At the same time, Haworth and Meaburn argue that the identification of sets of genes associated with learning disabilities is not likely to have a “direct impact on teachers in the classroom confronted with a particular child with a learning problem” (178). However, they do think that the prediction of genetic risk, for example, with a micro-array for learning disorders, will have important consequences for the (response to) treatment and prevention of learning disorders, as part of the “personalized medicine” movement. In addition, they consider that greater insights into environmental factors could also contribute to knowledge about individualized learning for all children.

Interviews and focus group discussion

Similar to the literature review, most of guiding visions expressed during the interviews and focus group were related to health care. In addition to understanding the biological pathways leading to disease, visions of researchers focused on the aspects of diagnostics, therapy and prevention of learning disorders and mental illness. Participants commonly argued that knowledge of genomics could improve both the efficiency and effectiveness of the health and education system. In the future, diagnostic tests of the genome are expected to take place in less specialized settings, such as the clinic. Their vision of a desirable future involved screening individuals very early in life so that genetic information is accessible at all times. According to participants, this could make the health-care system more efficient because it will become faster and easier to diagnose a patient with the genotype at hand, and more effective because the correct treatment can immediately be started and the effect of therapy can be predicted. The different guiding visions expressed by the researchers have been summarized in Table 2.

Table 2. Overview of future visions expressed by neurogenomics scientists.

Possible applications
(1) Understanding of brain function: participants envisioned that by 2040 there will be more understanding of complex diseases and of individual variation in genome, epigenome and environment influences on specific functional networks in the brain. Some participants hoped that this knowledge would lead to an improved definition of disorders
(2) Antenatal prevention: participants expect that scientists will be able to determine the origin of 90% of cases of mental retardation by 2040, making it possible to better advise families about the probability of disorders. It would be desirable if by that time families would have prevention options available to them other than abortion, such as presymptomatic (gene) therapy in the womb
(3) Postnatal prevention: knowledge of the genome could play an important role in the prevention of neurological diseases, but also in the optimizing of brain development. A tailored lifestyle advice, based on the genome and epigenome, would be “a dream” for some participating researchers. Genome sequencing, before or at birth, could provide information on the kind of nutrition, sports or education from which an individual would benefit
(4) Diagnosis and treatment: according to participants, a national genomics database would mean faster diagnosis because doctors would have more information about the probability of disorders in individual patients. Furthermore, they expect treatment to be more efficient because doctors know which medicine would be effective for individual patients

In the next sections, we relate the researchers' guiding visions to the second-order discourse of underlying theories and appreciative systems. These underlying theories can be clearly linked to the participants' diverse scientific backgrounds and empirical practice, as predicted by Grin and Van der Graaf (1996). Genomics research for education involves different types of research methods, including quantitative genetic methods, typically employing general population twin and adoption designs, and molecular genetics, which is useful to identify which genes are associated with which behavioral outcomes (Petrill and Justice 2007). Within these different fields, scientists from different disciplines are involved (e.g. molecular genetics brings together molecular biologists and specialists in bioinformatics) (Torgersen 2009). The participants of this study describe themselves as molecular biologists, statistical geneticists, neurologists, psychologists or researchers of early childhood, using research approaches ranging from mouse genetic strategies to genome-wide association scans (GWAS) and next-generation sequencing, to hypothesis-based genetic research with children with learning difficulties.

Links between visions and empirical approach

Researchers from different disciplines described a great variety of problems that they wished to solve with their research, most of which were consistent with their specific empirical approach. Differences in research strategy not only influence the way researchers work from day to day, but also the kind of questions they pose. For example, when asked about future applications of their research, researchers in mouse genetics considered that translation of their mouse models to human phenotypes would generate greater understanding of human brain function. In this way, some participants stayed very close to their field of research with their applications taking the limited step from animal to human.

In contrast, the future visions of researchers using a GWAS approach or other forms of molecular genetics involved a tailor-made lifestyle advice based on information about the individual's genomic and environmental status. This would require both an enormous amount of data and insights into the extremely complex interaction between genes and environment. These scientists' future vision reflects their inductive empirical approach, namely the exploration of enormous amounts of data in order to find statistically relevant relations between genes and phenotypes.

According to participants who are undertaking deductive, hypothesis-driven research into a single gene, it is neither desirable nor feasible for genomics research to contribute to improved learning in the general population. This group did not support the potential use of genotypes to identify learning phenotypes for the fine-tuning of educational strategies in the classroom. According to a researcher in early childhood involved in genetic research, it does not really matter which type of education most children receive because they generally continue to learn regardless. Moreover, she questioned the value of identifying learning phenotypes because she argued that most children's learning could not be made more effective or efficient by such an approach. From this point of view, genetic research could only provide benefits for very specific groups that experience problems with the current education system.

These examples illustrate that the type of empirical approach used by the researchers, often grounded in their disciplinary background, influences the possible applications identified, as well as the distance of these ideas from the practice of education. Researchers also stay close to their area of methodological expertise during the identification of barriers in the translation of genomic findings into educational practice. During the focus group, the participants, for example, discussed the possibility of individual lifestyle advice based on genotype. They considered that the greatest challenge to this advice is gaining more insight into the influence of environmental factors. However, the participants had different perceptions of the location of this knowledge gap, leading to different approaches. One approach proposed involved animal testing which can be used to relate variation in gene expression and environment to behavior. Another involved large-scale, longitudinal research measuring the environment of a large group of subjects over their lifespan. A third proposed approach focused on the effect of different environmental micro-interventions on children with regard to one specific gene. These three proposed approaches, and the problem definitions, overarching theories and appreciative systems from which they originate are clearly embedded in the disciplinary backgrounds and the consequent day-to-day research practices of the participants, offering very different solutions to different problems.

Different understanding of key concepts

We have seen that future applications of neurogenomics in the field of education envisioned by researchers are strongly influenced by their disciplinary background and empirical approach. In our analysis of the interviews and focus groups, we also found that the concepts of phenotype, intelligence and environment were central to participants' understanding of genomics in relation to education. Here, we aim to gain insights into researchers' guiding visions by analyzing how participants make sense of these three concepts which they construct on the basis of their overarching theories and appreciative systems.

The definition of phenotypes

In both the interviews and the focus group, participating researchers appeared to be divided in their opinions of how disorders should be defined, namely on the basis of genotype or phenotype. For example, many heterogeneous disorders of the brain are considered to be related to a spectrum of phenotypes (Sabb et al. 2009). One participant involved in hypothesis-based research into biological mechanisms underlying ADHD suggested that all people can be found somewhere on that spectrum of phenotypes, making it difficult to classify people into discrete categories related to their learning abilities or disorders:

It is more interesting to look at genetic factors than at these kind of categories. I thought it was fascinating to see that there are tests that are very suitable to measure ADHD but unsuitable to predict what will happen in education. […]

From this perspective, every individual has his or her own unique “composition” of genes, resulting from the complex interactions between genes, and between genes and environment. Research at the level of the genome would make it possible to look at individual variation and design tailor-made interventions. The same participant criticized the new Diagnostic and Statistical Manual of Mental Disorders' (DSM-5) focus on the symptoms of disorders rather than their etiology, arguing that it makes effective treatment more difficult:

The DSM is purposely ignoring the cause of these factors. It is really saying let's just deal with the symptoms as they present themselves to us and let's not look at the causes because that is too complicated. I take a different view. I think you would be better off looking at causes, because if the causes aren't there, well…. Treatment with methylphenidate is based on the causes, that's the whole rationale behind it.

This participant is of the opinion that the genetic factors contributing to the disorder should be the basis for the treatment of disorder. He argues that the DSM-5 reduces disorders to their phenotypes, not taking into account the complexity of their causes. The phenotypes of disorders should therefore be redefined on the basis of their genotype. However, other participants argued that this participant reduces the disorder to its genetic origin, eliminating the phenotype.

In contrast, other participants suggested that it was both possible and desirable to classify behavior into separate phenotypes on the basis of observable symptoms. As one interviewee, involved in GWAS studies of intelligence, commented:

In the end, it does not really matter whether it is genetic or not. Once you have the label, you know that you have it [the disorder]. It's really about the definition of subtypes. For example in ADHD: is someone of the passive or of the non-attentive type? You have to offer different forms of education to these children.

Here, the subtypes of spectrum disorders, listed in, for example, the DSM, are regarded as a basis for determining treatment and participants with this view sought the distinction between behavioral phenotypes in order to find the genes that contribute to these phenotypes. Indeed, much genetic research has been devoted to the matching of categories of disorders to predisposing genes (Holden 2008; Hariri 2009).

The above analysis also gives insight into the epistemic views of the researchers. It can be argued that the second group of participants, who wish to work from the phenotypes of the DSM to find out what genes are involved, have a more constructivist view of science than the first group who first want to gain more insight into the brain in order to re-evaluate current phenotypes. This latter view is illustrated further by the following interviewee:

There is a lot of variation in phenotype expression, possibly due to nurture, partly due to the influence of the rest of the genome. […] And so people with the same mutation can have a different phenotype. […] In autism there might also be a very strong component, you have a number of components, but not an unlimited number, that I do not believe in.

The researchers were encouraged to think into the future, but many of the visions they expressed dealt with answering basic scientific questions and gaining more understanding about the brain. Researchers wanted to gain understanding of the black box of the brain and discover its inner workings, the ultimate goal being insights into the mechanisms at work. Accordingly, they accept the value of mapping the genome, even though the processes that lead to the expression of genes are very complex.

Understanding of intelligence

Intelligence is a particular type of phenotype that we would like to discuss in more detail. On the one hand, similar to the ADHD example given above, some scientists involved in non-GWAS studies argue that the concept of “intelligence” needs to be reconstructed on the basis of biological factors, rather than vice versa. At the same time and reversely, multiple GWAS studies have been done to unravel the genetic factors underlying “intelligence”, measured as IQ, the scoring of which is categorized into sublevels. One participant involved in this kind of study explained that IQ is used because it is a reasonably standardized measure, widely used already. During an interview, this participant noted:

“We generally do research to find out why people become ill, be it psychiatric diseases, diabetes or cancer. But also why people differ in eye color, IQ, etcetera.”

[…]

Interviewer: How do you measure intelligence?

“At the moment we measure it simply with an IQ test. There have always been many criticisms to this, which is justified, because … what exactly is intelligence? I am always very pragmatic, so I call what the IQ test measures intelligence, and I take that with me in research. But of course, this is not completely true because intelligence is more than only an IQ-test. But that is very difficult to measure.”

This excerpt shows that for this scientist, IQ is one trait in a much longer list, of which some are more complex than others, and of which some would be categorized within the domain of health care, while others would not. The researcher is not involved in questions about what the traits entail exactly. Her interest lies with the genes themselves: the underlying mechanism, the puzzle of how the genome contributes to disorders and traits in general. This is in contrast with the interest of researchers for whom intelligence itself is the main focus of study and who wish to reconstruct the concept of intelligence.

In the excerpt from the GWAS researcher above, intelligence is viewed as a trait, which has a deterministic quality. A similar perception is visible in this vision described by a molecular geneticist during the focus group:

Participant 1: “I think that profiling for certain professions is not a bad idea at all. That is what we are doing now already. We advise people on the basis of diseases in the family. Like if your child has some kind of mutation. It's not a reason for abortion, but she will never be a top-athlete. She should avoid this and this, and she should avoid that and that.”

[…]

Participant 2: “Yes, but the point is, how far do you go?”

Participant 1: “Yes, if you say I want only good people and I want to prevent people with a defect [you go too far]. You can also point people with a defect in the right direction. Then you say to them that they can become very happy if they do this and this.”

This scientist does not consider selection for particular determining traits to be problematic because he does not make a value judgment about the trait itself. The ethical principle of equity is therefore an important part of this participant's reasoning. By paying attention to the differences between people, he argues that all people are given equal opportunity for happiness. As the excerpt shows, this vision was considered to be problematic by other participants.

In the scientific literature, intelligence is also a topic for debate because of the way it is measured but also because of its social consequences (Hunt 2011, p.1). According to Gray and Thompson (2004), the misconception that differences in intelligence imply a trait-like quality, independent of context and impervious to change, leads to some uneasiness within society with regard to research into intelligence. They argue that this is based on the fallacy that the way things are (individual differences in intelligence) implies something about how it should be. Evidently, this issue has not been satisfactorily resolved as researchers themselves hold different understandings of the concept. This has implications for the kind of research considered desirable.

Perception of environment

Our last example concerns the understanding of the environment in relation to genes. When it comes to the relationship between genes and environments, all participating researchers are “interactionists” in some way. They acknowledge the interaction between genes and the environment in the development of traits. However, the majority of participants hold that every trait starts with a certain genetic basis that may be modulated by environmental factors. A typical way of expressing this position is the following quote:

I think that, for the largest part, it [IQ] is genetically determined. Like I said, the borders are determined by the genetic context, and the scope is determined by nurture.

The statement reflects a mainstream idea in genetics research that is found throughout the literature, namely that environment merely modifies the expression of genes, either strengthening or weakening their effect on individual's phenotypes (Pennington et al. 2009).

A few researchers are interactionists in a different way, adopting a more fluid and holistic view in which nature (“genes”) and nurture (“environment”) can be argued to be an extension of each other. For example, one of the researchers mentioned that he “had lost his distinction between environments and genes” because he could not pinpoint where genetic influence stops and the environment begins. This is consistent with the ideas of the philosopher Peter Goldie (2002), who has criticized the idea that for every characteristic there is a biological essence (the hard core) that might be superficially modulated (the soft skin) by environmental elements such as social and cultural factors, calling it the “avocado pear misconception”. Goldie argues that it makes no sense to disentangle what is hard and biological and what is soft and cultural. A trait-like intelligence should rather be seen as developmentally open or plastic.

Interestingly, this difference in the conceptual scheme of nature versus nurture also reappears in different research choices and ideas about future intervention opportunities. The researchers who were more focused on the biological hard core recommended research directions that were very gene-focused, whereas the more interactionist researchers proposed research on the modification of the environment, making the environment the focus of intervention. For example, in the future vision of the genomics-based lifestyle advice, individuals can modify their internal and external environment to their genomic status. This vision would entail that individuals might continuously monitor factors within their environment and modify their environment or behavior according to the results.

However, in practice, the aim of many types of research is to “extract” the environment from the gene, to isolate the environmental influence as a distinct determinant. This comes from a much more dichotomous view of nature and nurture. Genes cannot be taken out of the environment but, in genetically sensitive research designs, researchers can investigate to what extent the environment can account for the phenotypic difference between people with the same genotype in order to identify to what extent certain traits are hereditary. Typical genome-wide studies do not take the environment into account, although this is expected to be an important next step in unearthing the genetic contribution to multifactorial disorders (Hamza et al. 2011). These types of research are actually very gene-focused (although they also contribute to knowledge about the environment) as is demonstrated by the interest of researchers with this type of framing in the possibilities of gene therapy.

Discussion

The field of neurogenomics is developing rapidly, but not perceptibly in the direction of education. Most of the future visions identified in the literature either involve the understanding of the biological pathways that lead to brain function, or they are clinically oriented, focusing on the diagnosis, prevention, treatment and prognosis of learning disorders such as dyslexia, ADHD and autism. During the interviews and focus group discussion, we steered the participants toward applications in the field of education and encouraged them to think into the far future. This resulted in two different reactions. Some scientists stayed quite close to their own research setting and were very hesitant to formulate concrete applications. Others took a completely different approach, presenting a rather optimistic view of what neurogenomics might become. Some envisioned that neurogenomics research could serve as a panacea to all problems with a genetic component. Scientists, for example, hardly problematized the use of neurogenomics for mass screening in the context of dyslexia, ADHD and autism, even while more nuanced views of screening are commonly held. The World Health Organization, for instance, emphasizes various ethical issues such as the balancing of risks for false negatives and false positives (Wilson and Jungner 1968; Andermann et al. 2008).

In this exploratory study, we aimed to reflect on the guiding visions of neurogenomics scientists for the field of education. However, we have not found clear visions that were shared by the participants. When looking into the second-order notions of the participants, we have seen that they have different overarching theories and appreciative systems, mostly relating to the disciplinary field of which they are part. Disciplinary background, empirical practice and understanding of key concepts (such as “phenotype”, “intelligence”, and “environment”), all of which are strongly related, influence the choices made by researchers. For example, participants using a hypothesis-based approach have a different epistemic view and think differently about the brain from those involved in GWAS research, leading to different guiding visions. Similarly, researchers' perspectives regarding the interaction between genes and the environment have implications for the kind of applications they envision for their research.

The different viewpoints within genomics could be mutually enriching. However, interactions between disciplines do not frequently occur. In particular, the perspectives of researchers in the early childhood and genomics seem to be far apart. Many of the researchers stress the importance of interdisciplinary communication and cooperation. However, in order to communicate on the same level, it might be helpful to pay attention to the differences in empirical approach that lead to different conceptual and epistemological views, and the differences in moral appreciations. Researchers within different disciplines can have different ideas of what is possible, and what barriers have to be overcome. Even though different disciplines can study the same object, it is likely that the understanding of that object will continuously need to be negotiated (O'Day et al. 2002).

We have conducted this study in an early phase of neurogenomics development because technologies become entrenched and difficult to steer in later phases (Collingridge 1981, 11). However, this makes prediction of the social consequences of technologies more difficult. A challenge we faced is the chaotic and unorganized nature of emerging technologies due to the absence of an “innovation system” (i.e. well-defined actors, relations and institutions) (van Merkerk and Smits 2007). Given that genomics research with possible implications for education takes place within many different disciplines, it was difficult to identify which research and researchers are part of the emerging field of “neurogenomics and education” and should therefore be included in this research. A lack of shared guiding visions and the differences in understanding of key concepts between neurogenomics researchers are indicators that the field we have tried to map out does not yet exist in the Netherlands. It is likely that this has also influenced our response rate, especially with regard to the participation in the focus group, because we think that some researchers did not consider the domain of education to be a priority in their research.

Based on the results of this exploratory study we have the following recommendations for future research and research policy that could contribute to the socially robust embedding of neurogenomics. First, it is clear that before we can effectively discuss potential opportunities and concerns, it is necessary to negotiate shared frames of meaning within the interdisciplinary scientific field of neurogenomics. A second step in stimulating desirable applications and preventing negative consequences of neurogenomics research would then be the further deliberation on potential developments in the (far) future, both within the Netherlands and elsewhere. Furthermore, in this study, we aimed to go beyond the application of neurogenomics in the medical field, but it is clear that more research is necessary into ways to deal with the incremental shift of medical technologies into other fields. Finally, given that a value-sensitive, inclusive approach is considered to be an important part of responsible research and innovation (Owen, Macnaghten, and Stilgoe 2012), we would also recommend further research into how alternative perspectives on emerging fields, such as neurogenomics, can be included in research agenda setting. One possibility could be to start discussion about the application of neurogenomics in the domain of education from the societal side, by gaining insight into potential end users' guiding visions for education and investigating in which way neurogenomics could contribute to the realization of these desirable futures.

Acknowledgements

This article is the result of the research project Neurosciences in Dialogue, part of the Centre for Society and the Life Sciences in the Netherlands, funded by the Netherlands Genomics Initiative, and the thematic program MVI (Socially Responsible Innovation), of the Netherlands Organization for Scientific Research (NWO). The authors are grateful to two anonymous reviewers for their insightful comments and to Sarah Cummings for her assistance in editing the manuscript.