The Evolution of Fine Motor Control and Language: Are Handedness and Language Lateralization Related?
By: Suzanne Martens
Program: Research Master Brain & Cognitive Sciences Track: Cognitive Neuroscience
Institution: University of Amsterdam Student ID: 10194045
Word count abstract: 280 Word count thesis: 14158 Date: 04-03-2016
Supervisor: Dr. Reint Geuze Co-assessor: Dr. Jelle Zuidema
Index
Abstract p. 3
§1. Introduction p. 3 - 7
§2. Evolution of Hemispheric Specialization p. 7 - 9
§3. Fine Motor Control and Language in our Ancestors p. 9 - 22 §3. 1 Laterality of Fine Motor Control in Primates p. 9 - 15
§3. 2 Language Abilities in Primates p. 15 - 22
§4. Fine Motor Control and Language in Humans p. 22 - 27
§4. 1 Genetics and Development of Handedness and Language Lateralization p. 22 - 24 §4. 2 Relationships between Handedness and Language Ability p. 24 - 27 §5. Handedness and Language Lateralization: Shared Origins but p. 27 - 29 No Direct Relationship?
The Evolution of Fine Motor Control and Language: Are Handedness and Language Lateralization Related?
Abstract
Research shows that handedness is related to language lateralization. Gestural theory proposes that this relationship exists because lateralization of fine motor control caused language to become lateralized in the same hemisphere. Qualitative change theory on the other hand proposes that language evolved relatively abruptly in humans before handedness emerged. The emergence of Broca’s area for fine motor control of speech related muscles caused lateralization of all fine motor control in the same hemisphere according to this theory. However, neither theories take in to account that hemispheric specialization is a general principle of brain organization, present in many species and across many functions. General mechanism theory therefore proposes that lateralization of fine motor control and language are independent from each other. Studies investigating fine motor control in our ancestors provide evidence that handedness evolved before language, but the kind of handedness primates display differs from human handedness in ways that are problematic for gestural theory. Qualitative change theory is challenged by these results as well, and by the observation that precursors of language present in primates already are left lateralized. General mechanism theory is not contradicted by either type of studies. In humans, development of lateralization of language and fine motor control seems to be influenced by genetic factors quite strongly, but does not seem to be related to one another. Furthermore, lateralization of fine motor control only has small effects on language abilities during development and has no effect in adults. The relationship between lateralization of language and fine motor control therefore seems to have been overestimated before. General mechanism theory therefore gains more support from this thesis than gestural and qualitative change theory.
§1. Introduction
Handedness is a phenomenon that shapes our daily lives in very practical ways. Because about 70-90% of the population is right-handed (RH) (Kolb & Whishaw, 2009), our world has been optimized for the use of the right hand. For example, doors typically open in the direction that allows for the easiest manipulation with the right arm. Even our bathroom habits show a right hand bias: a roll of toilet paper will typically be found on the wall to the right of the toilet. About 10% of the population is left-handed (LH) however, and approximately 7.5% is ambidextrous (AH, meaning they have no strong preference for either hand or prefer a different hand for different tasks) (Kolb & Whishaw, 2009). For centuries, being non-right handed (NRH) has been strongly discouraged by influential institutions such as primary schools and religious organizations (Sato et. al., 2008). Even without these prejudices, being NRH still has disadvantages seeing as the environment has been adapted to fit the right-handed best. Assuming our population wide hand preference is a result of the evolutionary process of natural selection, it would be expected that the prevalence of NRH would decrease until none are left. However, the proportion of the population that is NRH1 has
1 Coren and Porac evualuated the prevalence of LH in 5000 years of historical works depicting unimanual labor such as writing and painting. An estimate of the prevalence of AH could therefore not be made. To the knowledge of the author no such estimate exists. The prevalence of LH however was shown to be +/- 10% for any time point in these 5000 years.
stayed approximately equal for at least 5000 years (Coren & Porac, 1977). The existence of NRH is therefore quite a mystery.
NRH has inspired a lot of research. Some of this research, like the aforementioned research on prevalence, has focused on hand preference (a person’s tendency to use one hand over the other). Others have focused on hand skill level (a person’s ability to use a given hand for a given task). Hand preference is typically assessed using questionnaires or simple observation of the hand a person chooses to use when performing a series of tasks. Hand skill is typically assessed using a task that requires fine control of the movement of the hands. For example, the time it takes subjects to put pins in a board with holes can be compared between hands. Both preference and skill could be considered part of the construct handedness, but are only moderately correlated (Todor & Doane, 1977). This means relationships between variables and hand preference might differ from
relationships between the same variables and hand skill level. It is therefore important to consider what has been investigated in a particular study: hand preference, skill or both.
A frequent finding in the literature on handedness is that NRH preference (more so than skill) is associated with developmental disorders affecting many cognitive functions at the same time, such as autism (Fein, Waterhouse, Lucci, Pennington & Humes, 1985; Soper et. al., 1986; Hauck & Dewey, 2001), schizophrenia (Sommer, Ramsey, Kahn, Aleman & Bouma, 2001; Delisi et. al., 2002) and ADHD (Rodriguez, 2010). NRH preference is more common in these groups compared to the general population (e.g. as much as 40% is NRH in the case of autism: Fein, Waterhouse, Lucci, Pennington & Humes, 1985) and higher degrees of AH (though typically not LH) are associated with a higher severity of symptoms (Soper et. al., 1986; Rodriguez et. al., 2010). Researchers have therefore suggested that NRH might be a result of brain damage early in development (Satz, 1973). Though this might be the cause of NRH in some cases, a vast majority of NRH people does not suffer from brain damage and developmental disorders (McManus, 1983). Modelling NRH as a direct result of brain damage is therefore too simplistic.
Other frequent findings in the literature on handedness are related to hemispheric
specialization. The typical brain is not organized symmetrically. Though the left and right hemisphere contain roughly the same structures, structures in the left hemisphere do not necessarily process the same information in the same way as their equivalent in the right hemisphere does. Instead, the left and right version of a certain area often process different aspects of the same type of information. This is called hemispheric specialization. Brain regions involved in general motor skills clearly demonstrate hemispheric specialization. The left motor cortices control the right side of the body and the right motor cortices the left (Kolb & Whishaw, 2009). Fine motor skills show similar hemispheric differences as well. The term fine motor skills refers to the ability to coordinate small muscle movements, like the ones required for writing. Research on both preference and skill shows that the hemisphere opposite to the dominant hand is more involved in and more adept at
producing movements that require fine motor control (Dassonville et. al. 1997; Volkmann et. al., 1998; Bernard, Taylor & Seidler, 2011).
The most salient asymmetry in the typical human brain is the organization of language abilities. Both language processing and production are left lateralized to such a degree that the areas associated with these functions do not have a functional equivalent in the right hemisphere
(Wernicke’s area for comprehension and Broca’s area for production, see Figure 1) (Kolb & Whishaw, 2009). This became clear for the first time when neuropsychologists discovered that most aphasia patients (people having trouble producing or understanding language as a result of brain damage)
suffer from left hemisphere damage in Broca’s or Wernicke’s areas. Damage to the same areas in the right hemisphere however does not typically produce aphasia symptoms (Carroll, 2007).
More evidence for left hemisphere language lateralization (LHLL) was provided after the development of the WADA test. The WADA test was developed to predict chances of success and of cognitive impairment of ablative neurosurgery in epilepsy patients. During the WADA test, sodium amobarbital is injected in to either the left or the right internal carotid artery to inhibit the
corresponding hemisphere. By comparing performance before and during the WADA test on cognitive tasks, it is possible to determine how much cognitive functions are affected by the inhibition (or surgical damaging) of one hemisphere (Meador & Loring, 1999). When testing the left hemisphere, people are typically unable to speak and have trouble interpreting speech from others. Right hemisphere inactivation on the other hand does not have these effects (Kolb & Whishaw, 2009), showing that at least in this patient group the right hemisphere is not essential for language related functions while the left is (this does not mean the right hemisphere does not contribute to language processing, but it’s clearly involved to a lesser degree). Finally, non-invasive methods like fMRI and behavioral tasks have been used to investigate language lateralization in the healthy population as well (Kolb & Whishaw, 2009). Hund-Georgiadis, Lex, Friederici and von Cramon (2002) for instance used a behavioral task outside of the fMRI scanner as well as in the scanner in a sample of healthy participants to investigate language lateralization. During the behavioral task (the dichotic listening task), subjects were presented with audio recordings of two different words in each ear simultaneously and asked to repeat what they heard. Most subjects repeat the word heard by the right ear more often, which indicates a left hemisphere dominance in language processing because the left hemisphere is responsible for processing input from the right ear. In the scanner, subjects
Figure 1. Typical language lateralization. In the left hemisphere Broca’s area and Wernicke’s are present, structures concerned with respectively language production and comprehension. The arcuate fasciculus is a bidirectional bundle of axons that directly connects Broca’s and Wernicke’s areas. The same areas and connections in the right hemisphere are not crucial for language abilities.
were asked to categorize words as verbs or nouns while their brain activity was measured. Here too a strong left hemisphere activation (in Broca’s area) was found, but no right hemisphere activiation. Both findings confirm that language abilities are typically left lateralized in healthy populations as well.
LHLL has proven to be more of a given for RH than NRH people. Using the methods
described earlier, it has become clear that both right hemisphere language lateralization (RHLL) and no language lateralization (NLL) are about eight times more common among people with NRH preference than those with RH preference. Of NRH, about 8-15% is RHLL and 14-15% is NLL. Of RH, only 2-6% shows either type of atypical language organization (Szaflarski et. al., 2002; Kolb & Whishaw, 2009). Furthermore, NRH people with typical LHLL show a decreased strength of this lateralization compared to right handers (Gonzalez & Goodale, 2009; Kolb & Whishaw, 2009; Vingerhoets et. al, 2012; Constanzo et. al., 2015).
This research suggests that handedness and language lateralization are related. This has inspired several theories in the field of evolutionary psychology that can be divided in to two classes. The first class, referred to here as gestural theory, suggests that lateralization of language and fine motor skills are related because language has evolved directly from non-verbal communication of primates (Carroll, 2007). Gestural theory assumes that gesturing requires a certain amount of fine motor control, and that this skill was already lateralized to some degree before the emergence of language. Language has some things in common with gestures: it is used for communication, and it requires a great deal of fine motor control (that of the mouth, vocal tract and facial muscles). This suggests that gesturing and language might share some neural machinery. The utilization of pre-existing structures and mechanisms is called exaptation and is known to occur frequently during evolution. Given the similarities between gesturing and language in required machinery and exaptation as a possible evolutionary mechanism, it makes sense to propose that language would develop in the same hemisphere as fine motor control (Carroll, 2007).
The second class, referred to here as qualitative change theory, suggests that language lateralization has driven the development of handedness instead of the other way around. This hypothesis is based on Chomsky’s notion that language is a too complex ability to have developed gradually in the short time span between the emergence of the Homo Sapiens and the first ancestor that didn’t possess language abilities (Carroll, 2007; Papadatou-Pastou, 2011; Bolhuis, Tattersall, Chomsky & Berwick, 2014). This theory assumes that language does not quantitatively differ from the systems of communication our ancestors used, but qualitatively. A genetic mutation or a rapid series of them would have enabled the brain to be wired in a way that allowed humans to merge two elements (such as “brown” and “horse”). This ability is theorized to not be present in ancestors and to underlie all our grammar abilities, advancing human language past the forms of
communication our ancestors used. Assuming that this sudden emergence of language occurred before the emergence of handedness, lateralization of fine motor control for the vocal tract, mouth and facial muscles could have created a bias to use the left hemisphere more for all types of fine motor control.
It remains unclear how language lateralization and handedness are related. Is one type of lateralization the result of the other, like the two theories suggest, or is there an underlying cause for both types of lateralization that makes them seem related without having direct influence on each other? In this article, the nature of the relationship between handedness and language lateralization will be further investigated.
To examine this relationship, evolution of lateralization in general needs to be examined in detail. The two theories described both assume lateralization of either language or fine motor control before the other, but do not explain why there is lateralization in the first place. Possibly, the evolution of hemispheric specialization in general can illuminate the relationship between different types of lateralization. The first paragraph of this thesis will therefore discuss evolution of
lateralization in general. In the second paragraph, primate handedness and language lateralization will be investigated to examine the evolution of fine motor control and language. Next, the
relationship between handedness and language lateralization in humans will be discussed. Finally, a general conclusion with regard to the mentioned theories will be drawn.
§2. Evolution of Hemispheric Specialization
As is evident from the existence of hemispheric specialization in language and fine motor control, the left and right hemisphere are not identical. Though the differences between
hemispheres are most obvious for language and fine motor control, these functions are not special cases. In emotional processing for instance the left hemisphere is biased towards processing emotions that are associated with taking action (e.g. affection and anger), while the right
hemisphere is associated with processing emotions associated with withdrawal from the stimulus (e.g. disgust and sadness) (Murphy, Nimmo-Smith & Lawrence, 2003; van Honk, & Schutter, 2006). In visual processing, the left hemisphere is more concerned with the details of a visual scene while the right hemisphere keeps track of global changes (Fink et. al., 1997; Rossion et. al., 2000).
Asymmetries like this have been found for most functions and modalities that have been studied extensively (e.g. memory, olfaction, attention etc., Kolb & Whishaw, 2009). Hemispheric
specialization therefore appears to be a general principle of brain organization in humans.
Hemispheric specialization used to be considered a uniquely human phenomenon because the first data showing it exists was Broca’s data on language lateralization. Because language is often considered a uniquely human ability, lateralization was proposed to have developed to optimize language function and therefore to only be present in humans (Harris, 1999; Hopkins, 2007). This idea has been refuted in many ways since these first discoveries in the late 19th century. As already mentioned, hemispheric specialization is not unique to language. Many systems that developed before language during evolution, such as the olfactory and visual systems, are organized according to this principle as well. Moreover, experiments have shown that many species without language abilities possess asymmetrically organized brains (Vallortigara & Rogers, 2005). The left
hemisphere/approach and right hemisphere/avoidance distinction in emotional processing seems especially wide spread in the animal kingdom. Toads, chicks and dunnarts are all more likely to flee from a predator when it is presented in the left visual field (right hemisphere) and attack prey (or a passive food source) that is presented in the right visual field (left hemisphere) (Vallortigara & Rogers, 2005).
The fact that hemispheric specialization is present for many functions and in species that are far removed from each other on the phylogenetic tree suggests that this way of organizing the brain evolved early. Many hypotheses have tried to provide specific selection pressures that might explain the selection of hemispheric specialization. For example, the capacity hypothesis states that from a neuroanatomical point of view it’s redundant to have two identical hemispheres. Allowing for non-symmetrical brain development leaves space to develop more different functions, which is beneficial
and could therefore have driven the selection for hemispheric specialization (Vallortigara & Rogers, 2005). Inhibition models take a more cognitive approach. They state that having a dominant hemisphere for a certain process can be beneficial because it enables that hemisphere to inhibit inappropriate responses from the other (Allen, 1983; Vallortigara & Rogers, 2005). Finally, models that don’t propose a dominant hemisphere state that having two hemispheres work on the same information in a different way enables organisms to multitask, which is a highly adaptive skill (Allen, 1983).
These theories sound plausible, but are they likely to be true? A common mistake in
evolutionary reasoning is to belief that because of natural selection every trait an organism has must exist because it’s somehow adaptive. Some traits are simply byproducts of other traits that are adaptive. A classic example is the color of blood: blood is red because of hemoglobin, the molecule that very adaptively allows blood to transport oxygen, not because the redness itself provides some advantage. Natural selection is also not the only mechanism that plays a role in evolution. Other processes, such as genetic drift, can cause a non-adaptive trait to become prominent in a
population. It has therefore also been proposed that hemispheric specialization is not an adaptive trait but one that evolved by chance (Hopkins & Cantalupo, 2008). The fact that hemispheric lateralization also has theoretical disadvantages supports this idea. For example, unilateral brain damage in a lateralized brain causes loss of function, whereas such damage in a symmetrical brain does not.
It is difficult to find evidence for any of the specific hypotheses proposing natural selection of hemispheric specialization, especially because they are not mutually exclusive. There is evidence however that hemispheric specialization provides advantages and is therefore unlikely to have evolved by chance2. Indirect evidence has already briefly been addressed in the introduction: weaker lateralization (at least when it comes to motor skills) is associated with developmental disorders like ADHD, autism and schizophrenia. This implies that hemispheric specialization is important for our cognitive functioning in general. More direct evidence was provided by Rogers (2000). Rogers directly compared fitness of chicks with lateralized visual systems and chicks without this
lateralization. Normal chicks are right lateralized for predator responses. Lateralization of the visual system already starts in the egg, given that the egg is exposed to light. By depriving chicks of light during this period, Rogers created 17 non-lateralized chicks to compare against 19 normally
developed ones. These chicks were tested on their ability to detect a predator while picking grain in an apparatus that allowed for presentation of an image of a predator to one visual field. The groups showed no difference in ability to detect the predator (indicated by the amount of picks they made during exposure to the predator) in the right visual field (left hemisphere), but the lateralized chicks showed a clear advantage when the predator was in the left visual field (right hemisphere).
Lateralization for predator responses therefore provides an advantage for chicks about half of the time (assuming predators approach from the left or right side at random). These kinds of results imply that the mechanism of hemispheric is adaptive, and therefore likely to have evolved because of natural selection.
A similar debate can be conducted over the peculiar tendency of the side of lateralization to be a population wide phenomenon. In all examples so far, lateralization of a type of behavior has
2 There still remains a possibility that initially, hemispheric specialization in a primitive form evolved by
mechanisms other than natural selection. Adaptive properties could have evolved afterwards through exaptation. This is basically an untestable hypothesis however, and this subtle difference is not crucial to the subject at hand. It will therefore not be discussed here.
been found in the same hemisphere for the majority of the species (just like human handedness and language, which are both left lateralized in about 90% of the whole population). The theoretical advantages of hemispheric specialization do not depend on this property though. If 50% of the population would be left lateralized for an ability or behavior and 50% of the population right lateralized, hemispheric specialization would still enable the brain of each individual to store more different functions, to inhibit inappropriate responses from the other hemisphere or to multitask. In theory, having a population wide bias for lateralization side for predator responses could even be maladaptive (Vallortigara & Rogers, 2005). If 90% of an animal species responds faster to predators when they approach from the left side instead of the right, predators might learn to attack from the right side and kill more animals that are lateralized for predator responses in the same hemisphere. Here the same questions arise again: is a population wide side preference of lateralization an adaptive trait that is selected for, or does it develop by chance (even if it’s potentially maladaptive)? The most commonly mentioned potential benefit of having a population wide preference for side is that cooperation between individuals of the same species is easier when they behave in similar ways (Vallortigara & Rogers, 2005). With regard to handedness, this can easily be imagined. Demonstrating how to use a tool is easier when the student would use the same hand to manipulate it themselves (movements don’t have to be mirrored that way). With other abilities, it is not so clear why same side lateralization would enhance cooperation. The side of language lateralization of an individual for example can only be discovered under highly artificial circumstances like the dichotic listening task. How often would individuals of a species be confronted with such situations in real life, and how often would these situations require cooperation? Probably close to never. If the need for cooperation would cause population wide preferences for side, one would expect that only abilities that clearly behave differently under natural circumstances when lateralized one way or the other would show a population wide side preference. The conclusions of this debate are not clear. Possibly, a population wide preference for side develops independently from hemispheric
specialization. This implies that some behaviors can be lateralized for one side at the level of the individual but not at the population level without undermining the benefits of hemispheric specialization in general.
This paragraph has shown that even though it’s uncertain what environmental influences caused natural selection of hemispheric specialization in the first place, it is clear that it has advantages and is not limited to language, handedness or humans. This mechanism is not
undermined when there is no population wide preference for side of lateralization, but a population wide preference does seem to be the default. Hemispheric specialization seems to be a general mechanism, and is unlikely to have developed in order to optimize one type of cognitive process. Possibly, handedness and language lateralization are independent results of the mechanism of hemispheric specialization.
§3. Fine Motor Control and Language in our Ancestors
Gestural theory states that lateralization of language is a result of earlier lateralization of gestural communication, while qualitative change theory proposes the opposite. As the previous paragraph explained, the lateralization of fine motor control and language could also be
independent from one another but rely on the same mechanism (an idea referred to as general mechanism theory from here on). Each theory would predict different observations with regard to the presence and lateralization of fine motor and language skills in our closest non-speaking
ancestors (for an overview see Table 1). First, it is not controversial that non-human primates are capable of executing precise hand and finger movements. None of the theories therefore challenge the presence of fine motor skills in primates. However, whether these skills are lateralized in primates is predicted differently by the three theories. Gestural theory states that RH in primates caused LHLL in humans, so it follows that this theory would predict that primates already show a degree of RH. For RH to be able to influence the direction of language lateralization on the population level, RH must also be present on the population level. Individual handedness would therefore not support gestural theory if the amount of RH primates is not higher than the amount of LH primates. Qualitative change theory states that handedness developed as a byproduct of the development of the fine motor control needed for speech and would therefore predict that our ancestors do not show a preference for one hand over the other either individually or on the population level. Finally, general mechanism theory predicts that any ability shows lateralization. Whether this laterality is population wide or only present in the individual is not relevant. Since it is clear that fine motor control is present in primates, this theory would predict that primates show laterality in this domain.
Unlike the presence of fine motor skills in primates, the presence of language abilities (either in the gestural or vocal domain) in primates is very controversial. Gestural theory would predict that primates only possess gestural but no vocal language abilities. This is because this theory views gestural communication systems as precursors of human language, and vocal language as an ability that evolved in humans from gestural language abilities. Gestural communication systems of
primates should therefore significantly resemble human language, but vocal communication systems should not. Moreover, the theory states that lateralization of gestural language abilities caused vocal language abilities to become lateralized in the same hemisphere. Therefore, gestural language abilities of primates should show some left hemisphere laterality. Qualitative change theory would predict that primates do not possess language skills in either modality that significantly resemble human language. Since qualitative change theory does not consider vocal or gestural behavior of primates to be precursors of language, lateralization of vocal or gestural behavior in the left hemisphere would be hard to explain for this theory. Qualitative change theory would therefore predict that the communication skills primates display do not show lateralization. Finally, general mechanism theory does not predict whether language abilities are present in primates. However if they are present, they should also be lateralized (at least of the level of the individual). To examine the three theories, these predictions on fine motor control and language abilities in our ancestors will be investigated in this paragraph.
§3. 1 Laterality of Fine Motor Control in Primates
Laterality of fine motor control in primates has been a research topic of interest for a long time. In 1953, Warren already showed in a simple experiment that required 84 rhesus monkeys to reach for food that 63% of the monkeys used the same hand to pick up food on 80% of all trials. Half of the monkeys with such a strong preference used their left hand, the other half their right. This research suggests that most rhesus monkeys have a hand preference on the level of the individual. On the sample level however, LH and RH were equally common. When assuming lateralization of fine motor control affected lateralization of language, this finding is unexpected. Lateralization of fine motor control could only have caused LHLL on the population level if fine motor control itself was already lateralized on the population level to some degree before the emergence of language. Therefore, this finding actually seems to contradict gestural theory. Absence of population level
handedness does not contradict qualitative change theory or general mechanism theory.
Warren’s research was exploratory in nature and therefore didn’t address some issues. Most importantly, Warren (1953) only tested one species of primate. Primate species differ greatly from one another in many aspects. Important aspects in relation to our research question are those that enable us to order primates from “less similar” to “more similar” to humans. For example, the phylogenetic distance of the species to humans (see Figure 2 for an overview), the degree of social behaviors a species displays and general intelligence can all be important for our expectations of how much a primate species should reflect the human condition. An overview of handedness among different species is therefore needed.
Since Warren’s experiment, many studies have tried to examine handedness in primates in different ways and different species. Papademetriou, Sheu and Michel (2005) performed a meta-analysis over 118 studies of primate handedness published since 1987. In this meta-analysis, they investigated the proportion of AH, LH and RH monkeys found over all studies investigating one species of primate using a simple reaching task. Individual hand preference was determined by calculating z-scores based on the amount of trials one particular hand was used divided by the total amount of trials. If an individual primate had a z-score of lower than -1.96 or higher than 1.963, they were classified as LH and RH respectively. Primates with z-scores in between were considered AH. The data revealed that all species of primate show a certain degree of individual lateralization. The proportion of AH across all studies for prosimians, new world monkeys, old world monkeys and apes were respectively 12%, 19%, 33% and 23%. Five of the 14 species investigated showed a significant sample wide preference for one hand over the others. Of lemurs (prosimians) 58% were LH, 31% RH and 11% AH. Of tamarins (new world monkeys) 18% were LH, 76% RH and 6% AH.
Table 1. Comparison of the Predictions of the Three Main Theories on whether Lateralization of Handedness and Fine Motor Control are Present on the Individual and/or Population Level in Non-Human Primates.
Gestural theory Qualitative change theory
General mechanism theory Present Lateralized
(IL* or PL**)
Present Lateralized Present Lateralized
Fine motor control Yes Yes (PL) Yes No Yes Yes (IL or PL)
Gestural language abilities Yes Yes (PL) No No No prediction If present, yes (IL or PL) Vocal language abilities No No No No No prediction If present, yes (IL or PL) * IL stands for lateralization on the individual level
** PL stands for lateralization on population level
3 These values are typically used when conducting comparative tests like t-tests. In this case, a score of -1.96 or
lower and 1.96 or higher means that the chance of observing this particular proportion when the primate is actually choosing a hand at random is lower than 5%.
Of rhesus and Japanese macaques (old world monkeys) 58% and 52% were LH, 11% and 21% RH and 31% and 27% AH, respectively. Of apes, only chimpanzees showed a sample wide preference with 27% LH, 41% RH and 32% AH. These results show that individual lateralization is present among all primates, but that population wide preferences are rare. Even among species exhibiting population wide preference, the prevalence of AH is much higher than among humans. No upward trend is apparent in the data either. If anything, the primates further removed from humans seem to exhibit more lateralization on both the individual and sample level. Of our two closest ancestors (the chimpanzee and bonobo are equally far removed from humans) only the chimpanzee exhibits a sample wide preference for the right hand. These data are therefore inconclusive. Though it is clear that some individual lateralization of fine motor control is widespread among primates, it is still unclear whether population wide preferences exist.
It’s important to note that both Warren (1953) and the meta-analysis of Papademetriou, Sheu and Michel (2005) did not address a methodological issue. The test apparatus used by Warren allowed the monkeys to choose their posture themselves when taking the food. However, a
monkey’s hand preference for keeping balance by putting a third paw on the floor might override their hand preference for precision grasping. Laterality of fine motor control might be more obvious when monkeys stand on their hind legs. Papademetriou, Sheu and Michel (2005) purposefully excluded studies that used bipedal tasks for the sake of interpretability. If this
Figure 2. Simplified phylogenetic tree of primates. The tree does not include all species of primate. The species closest to humans according to shared ancestors are the apes, followed by the old world monkeys, new world monkeys and prosimians. Within the class “apes” chimpanzees and bonobos are our closest relatives, followed by the gorilla, orangutan and gibbon.
methodological issue is important, their meta-analysis might have given a biased view on primate handedness.
Braccini, Lambeth, Schapiro and Fitch (2010) addressed this issue by testing 46 chimpanzees in three different conditions. The hand the chimps used to manipulate a tool they needed for getting food out of a tube suspended above the head was recorded while sitting down, standing up near a wall they could use for support and standing up without support. Next, the chimps were divided in to LH, RH and AH groups using the same method as Papademetriou, Sheu and Michel (2005). The data showed that the amount of AH chimps decreased with an increase in bipedal posture. In the sitting condition, 48% of the chimps were classified as AH compared to 22% and 9% in the bipedal with and without support conditions. However in none of the conditions one type of laterality was more common than the other. These results show that testing in bipedal posture increases laterality in primates. However, despite this increased laterality, no sample wide preference for side was found.
More studies have found that a bipedal stance increases laterality in primate handedness studies. Hashimoto, Yamazaki and Iriki (2013) showed laterality increased under bipedal conditions in 12 marmosets (new world monkeys). Blois-Heulin, Bernard and Bec (2007) found a similar effect in 9 mangabeys (old world monkeys). Finally, Larson, Dodson and Ward (1989) demonstrated that posture also affected laterality in 10 bushbabies (prosimians). Furthermore, Larson, Dodson and Ward (1989) found a preference in this sample for the left hand. In total, three of the four studies found no preference on the sample level for one side over the other. These studies show that a tendency to display more lateralized behavior in bipedal posture seems to be present in all types of primate. Consequently, it can be argued that it needs to be included in research to avoid
underestimation of laterality. However, it does not seem to increase the odds of finding sample wide preferences.
The conclusion that no population wide preferences for side are present in primates cannot be drawn from the mentioned studies. First, the evidence is mixed. Second, the studies all used small to very sample sizes. It’s likely that they lacked the power to detect differences in the proportion LH and RH monkeys even if they did exist. Ideally, another meta-analysis on population wide
preferences would be performed using studies that included a bipedal stance in their design. Unfortunately, to the knowledge of the author no such meta-analysis is available.
The main benefit of the bipedal research design is that neither of the hands is involved in maintaining posture. The design has a large drawback too though, as it demands unnatural postures from the primates. Tasks that allow the primates to sit but also demand that both hands are
occupied might reflect handedness of primates in a more natural way. Picking up food is typically a unimanual task: the primate only needs one hand to complete it. The task can easily be modified in to a bimanual task however. For example, primates could be presented with an object with food in it. One hand would be needed to steady the object (which is not considered fine motor control) and the other would be needed to pry the food out of it (this is considered fine motor control). Blois-Heulin et. al. (2006) have shown that this design has the same benefits as bipedal designs. They tested 11 mangabeys tested for hand preference in different postures and while using one or both hands to complete tasks. To test for posture, the primates were confronted with conditions similar to the ones used by Braccini, Lambeth, Schapiro and Fitch (2010). To test for the effect of uni- and bimanual tasks, they used a simple reaching task like Warren’s (1953) and a puzzle task. In the puzzle task, the primates were confronted with a closed box with food in it. To get the food, they needed to
keep the lid open with one hand while prying out the food with the other. The results showed that the strength of lateralization as well as the amount of monkeys that showed lateralized behavior was higher in the bipedal and bimanual conditions compared to the sitting and unimanual conditions. Again, no preference for one hand over the other on sample level was found. Other studies with similar sample sizes have replicated the usefulness of bimanual tasks (Zhao, Gao & Li, 2010;
Canteloup, Vauclair & Meunier, 2013; Bardo, Pouydebat & Meunier, 2015). Zhao, Gao and Li (2010) and Bardo, Pouydebat and Meunier (2015) found sample wide preferences: a left hand preference in snub-nosed monkeys (old world monkeys) and a right hand preference in bonobos, respectively. One study uses bimanual tasks to investigate handedness in primates in moderate to large samples. Hopkins et. al. (2011) tested 536 chimpanzees, 76 gorillas, 118 bonobos and 47 orangutans using the bimanual tube task, which entails simply providing the primates with a tube with food in it. Essentially, this study is a replication study of Hopkins et. al. (2003). Data of the older study were presented again and merged with the new data for the analyses. Individual hand preference was determined using the same method as Papademetriou, Sheu and Michel (2005). The results showed that chimpanzees and gorillas displayed a significant RH and orangutans a LH preference. Bonobos showed no sample preference. In all three species showing a sample preference, 21-24% was AH. In chimps and gorillas, RH (50% and 54%) was about twice as common as LH (29% and 22%). In
orangutans, LH (57%) was almost three times more common than RH (19%). These results suggest that at least in apes, population wide preferences are present (though weaker than in humans). If these results are generalizable, this would support gestural theory.
A meta-analysis like Papademetriou, Sheu and Michel’s (2005) does not exist for bimanual (or bipedal) data. However Meguerditchian, Vauclair and Hopkins (2013) reviewed many articles on primate handedness that used a bimanual task to test their subjects. They found 11 articles that reported a sample wide RH bias and 7 articles reporting a sample wide LH bias in species of old world monkeys and apes (including Hopkins et. al., 2011). However they also found 10 articles reporting no sample wide preferences or restrictions on sample wide preferences (such as that only females showed a preference). This might not sound convincing at first, but some interesting differences with articles on unimanual tasks are present. First, of all 118 articles Papademetriou, Sheu and Michel (2005) reanalyzed, 47% originally reported finding no sample level biases for one hand over the other. For the bimanual case, only 36% of all articles used reported no sample bias (this 36% includes the articles with restricted sample wide preferences). Second, for every family of primates, Papademetriou, Sheu and Michel (2005) found multiple articles reporting conflicting results (e.g. 15 articles on new world monkeys reported RH bias, but 9 articles reported LH bias). Interestingly, all species that were reported showing a RH bias using bimanual tasks were never reported to show a LH bias, and vice versa. Though the amount of studies reporting no or restricted sample biases for fine motor control is still quite high, this review makes clear that bimanual tasks reveal handedness more often and in a more consistent way than unimanual tasks do. This indicates again that the tasks used to measure handedness matter, and gives us more confidence in the existence of population wide preferences in old world monkeys and apes. However, Meguerditchian, Vauclair and Hopkins (2013) do not describe how they conducted their literature search and did not include some articles on bimanual tasks described here earlier that did not find sample biases. Furthermore, for every species discussed only between one and four studies were mentioned, possibly because more are not available yet. For these reasons, the review doesn’t provide us with conclusive evidence.
This paragraph has provided extensive evidence for the existence of lateralized fine motor control in primates (for a summary, see Table 2). Clearly, our ancestors possess handedness on the level of the individual. The evidence for a population wide preference for the right hand is rather weak though, even in our closest ancestors. However, some methodological issues were shown to matter. These issues were present in the majority of older research and solutions for these matters have not been exhaustively investigated yet. Since there are some indications that population wide preferences can be found in a more reliable way using these solutions, it’s plausible that some primate species do possess a population wide preference for handedness (though probably weaker than the population wide preference of humans). It is unlikely that lateralization of fine motor control could have caused LHLL on a population level if there is no preference for side on the
population level for fine motor control itself. Therefore, studies on primate handedness provide only weak support for gestural theory at this time. Qualitative change theory on the other hand has been contradicted by every study described, since this theory does not allow individual handedness. General mechanism theory is not necessarily supported by this paragraph, but is not contradicted by any study either.
§3. 2 Language Abilities in Primates
Now that it has become clear that primates possess a form of handedness, language abilities need to be examined to fully investigate the three theories. This is difficult because there is no consensus on the definition of language. Some definitions are quite broad and focus on the function of language. Basically, these definitions treat all systems of voluntary communication (whether they are expressed gesturally, vocally, or otherwise) as language (Carroll, 2007). Other definitions are based on the differences between human communication and animal (not primates per se) communication and are therefore much more restrictive. These definitions demand grammar, infinity of possible expressions, and arbitrariness of symbols (meaning that symbols are acquired through social learning, not by genetic code) from communication systems before they count as language. If a system does not comply with one of these demands (instead of all of them), than that’s already enough to disregard it as a language system (Carroll, 2007).
As explained before (see Table 1), gestural theory would predict that primates show gestural but no vocal language abilities, whereas qualitative change theory would predict primates show neither type of language ability. But what does it mean to have language abilities? Since it proposes that the merge ability is responsible for the emergence of grammar, qualitative change theory inherently makes use of the strict definition of language (Bolhuis, Tattersall, Chomsky, & Berwick, 2014). Under the strict definition of language, an absence of the merge and grammar abilities, the ability to arbitrarily assign meaning to symbols, or the ability to produce an infinite amount of expressions would indicate that primates do not possess language abilities. However, the strict definition of language is based on differences between human language and animal communication. Using this definition encourages an anthropomorphical approach of the problem that almost ensures primate communication cannot satisfy the criteria. It is not possible to search for meaningful
differences or similarities between human and primate communication while demanding perfect equality at the same time. In this paper, the absence of one ability will therefore not be interpreted as evidence that no language abilities are present at all. Instead, language ability will be seen more as a continuous variable and evidence for the presence of any part of the strict definition of language will be interpreted as evidence for the presence of language abilities in primates.
Language abilities of primates have been investigated intensively. Some research teams have investigated gestural abilities by trying to teach captive apes a form of sign language (primates lack the physical machinery to produce speech sounds and can therefore never be taught to speak, Furness, 1916). Most of these projects were moderately successful. It is clear that learning a
language does not come naturally to apes and is not naturally used after acquisition to communicate with other apes (Tomasello, 2007), but extensive training does enable them to communicate on unexpected levels with humans. Chimpanzee and gorilla subjects raised with sign language (or another kind of non-verbal language) from birth can learn to understand and produce about 100-300 different signs and typically respond correctly when asked to retrieve an item, even when asked in a novel way (Seidenberg & Petitto, 1979). This shows that primates are able to learn what arbitrary signs refer to. This is an important condition for language learning and is part of the strict definition of language described earlier. The gestural language abilities of apes do not seem to extent much further than this though. Apes make many novel combinations, but most of these novel
combinations are meaningless. This behavior suggests that apes produce combinations more or less randomly and produce meaningful combinations by chance (Seidenberg & Petitto, 1979). It is therefore doubtful that apes are able to combine elements to form new meaningful elements. Indicators of an ability to use grammar, like word order or recursion, are also usually not observed in the utterances of apes (Seidenberg & Petitto, 1979). Some apes do seem to understand grammar, even if they can’t produce it themselves. Bonobo Kanzi is able to correctly respond far above chance level to question like “Can you make the snake bite the doggie?”. In sentences like these it would not be clear who is supposed to bite who without an understanding of grammar (word order in
particular) (Savage-Rumbaughm et. al. 1993). These results suggest that some of the neural machinery needed for gestural language abilities is already present in our closest ancestors. However, this neural machinery is not spontaneously used for language and no (or very weak) evidence has been provided so far for the presence of the specific abilities to merge, use grammar and produce infinite amounts of meaningful expression in apes.
Learning a type of human language is of course a highly unnatural activity for apes. When using their natural communication systems, primates might behave in different ways than while using their acquired human language. Researchers have investigated the communication systems apes and other primates typically use in the wild. Two systems are commonly observed in the wild: a system of verbal calls (called species specific calls) and a gestural system (Tomasello, 2007). Species specific calls are sounds that primates make in response to an evolutionary relevant stimulus. These stimuli are typically predators, new food sources, aggressive behavior and mating signals (Fedurek & Slocombe, 2011). The type of information conveyed by the calls can be quite specific. For example, different calls can refer to different types of predators (Zuberbühler, 2001). In chimpanzees, the call system is complex enough to refer to different kinds of food based on their desirability (Slocombe & Zuberbühler, 2006). Some old world monkeys (Arnold & Zuberbühler, 2006; Candiotti, Zuberbühler, & Lemasson, 2012) and chimpanzees (Crockford & Boesch, 2005) even seem able to make
meaningful combinations of calls. Social context can modify usage of calls to a certain extent, though typically only in adults (Egnor & Hauser, 2004).
These observations imply that vocal communication systems show some similarities to human language. Specific elements refer consistently to specific objects in the outside world, combinations can be made to express new information and to some degree calls are intentionally used to communicate with others. There are some important differences though. Repertoire size is in no way comparable to human language: even chimpanzees produce less than 40 unique calls
(McComb & Semple, 2005) and only 88 combinations (Crockford & Boesch, 2005). Even though chimpanzees are able to combine, their amount of expressions is finite. Moreover, primate calls are innate: primates raised in isolation produce them and react to them the same way non-isolated primates (Winter, Handley, Ploog, & Schott, 1973; Herzog & Hopf, 1984). When social learning adapts vocal calls, it adapts the context in which they are used and not, or barely, the sounds themselves (Lemasson, Ouattara, Petit, & Zuberbühler, 2011). Primate vocal communication therefore does not contain arbitrary symbols. Finally, combinations of calls do not seem to have grammatical structure other than call order (Arnold & Zuberbühler, 2012). This research shows that primate vocal communication systems are much simpler than human language. An important ability is present in primate vocal communication however that is not observed when examining primates who learned a human gestural language. Meaningful combinations can be formed, which indicates that primates do possess a basic form of merge ability. Since qualitative change theory states that only humans are supposed to have this ability, this contradicts the theory.
Gestural communication systems show some interesting differences with vocal
communication systems. First, the gestural repertoire of symbols is typically somewhat larger than the vocal repertoire. Chimpanzees for example display about 66 different gestures (Hobaiter & Byrne, 2011). Next, gestures are used more in social interactions (e.g. to ask another monkey to play) and are more sensitive to social context (Tomasello, 2007). Moreover, there is some evidence that shows gestures are arbitrary signals that are not transmitted genetically (Tomasello, 2007). For example, Laidre (2008) reported variability in gestures used by eight groups of mandrills (old world monkey) that either suggested social learning or invention of new gestures by individuals (both imply arbitrariness of symbols). One group displayed a gesture that was completely unique (Laidre, 2011). Finally, meaningful combinations of gestures are generally not observed, except maybe in bonobos (Pollick & De Waal, 2007). Other primates like gorillas (Genty & Byrne, 2010), chimpanzees (Liebal, Call, & Tomasello, 2004; Pollick & De Waal, 2007) and orangutans (Tempelmann & Liebal, 2012) often sign in sequences, but those sequences do not have a specific meaning like
combinations of calls do. Sequences are often repetitions of the same sign and seem to be produced mainly because the recipient of the message has failed to react to the signing primate. Possibly, the merge ability is only present in the vocal behavior of primates.
The data presented so far provides some evidence for the presence of the merge ability (albeit limited and possibly only present in vocal communication) and arbitrariness of symbols (oddly enough only present in the gestural domain) in primates and shows that vocal and gestural
communication systems are quite complex (for a summary, see Table 3). It would therefore be incorrect to infer that primates possess no form of language ability whatsoever, though it might be more appropriate to speak of precursors of language given the large differences with human language. The claim of qualitative change theory that a sudden emergence of language in humans caused handedness is therefore not supported by research on primate language abilities.
However a milder version of the theory could be formulated that would state that LHLL did not evolve until language started to require fine motor control of speech related muscles. This would still be consistent with the idea that it was the development of Broca’s area that caused a bias for fine motor control to be lateralized in the left hemisphere. Since the precursors of language discussed so far do not require extensive fine motor control of speech related muscles, this milder version of qualitative change theory would predict that the precursors of language present in primates do not show left hemisphere lateralization.
This new hypothesis could be investigated using vocal or gestural communication of primates. Data on lateralization of gestural communication is hard to interpret because handedness can be a confounding factor. If primates gesture more with their right hand, is that because they show a left hemisphere bias for communication or because they are RH in general? Meguerditchian, Vauclair and Hopkins (2010) found some indications that lateralization of social gesturing and fine motor control differs from one another. They observed the social gesturing behavior of 70
chimpanzees. The first difference the results show is that about 40-50% of the chimpanzees was AH for any type of social gesturing. Compared to the rates discussed in paragraph 3.1, this is quite a high rate of AH. This could be a result of the free nature of social communication. The chimps were not required to stand up or keep one hand occupied, factors which have been shown to increase laterality of fine motor control. Possibly, this relatively high rate of AH is only present because there were no restraints placed on the behavior of the chimps. A more striking difference in their data is the ratio RH/LH subjects. For all types of social gesturing observed, RH was about 5 to 12 times more common than LH (for one type of gesturing RH was 23 times more common than LH). For general fine motor control, research that reported a population wide preference only indicated that one type on handedness was about 2 to 4 times more common as the other. This research therefore suggests that social gesturing might be more strongly lateralized than fine motor control in general. Hopkins et. al. (2005) directly compared handedness for gestures and for fine motor control in 227 chimpanzees. Three measures of fine motor control were used: the tube task described earlier, a simple reaching task and a task requiring the chimps to use a tool to get food out of a tube. As a measure of handedness for social gesturing begging gestures towards human care-takers were used. For social gesturing, they found that 26% was AH. RH was about 3.7 times more common than LH. These results differ a lot from Meguerditchian, Vauclair and Hopkins’ (2010) findings and are more similar to the results on general fine motor control. Hopkins et. al. (2005) do not report the proportions of AH, LH and RH for the three measures of fine motor control, but do show using ANOVA that more chimps were RH for social gesturing than for the other measures. These results support the idea that chimps are more lateralized for social gesturing than general fine motor control.
Meguerditchian, Vauclair and Hopkins’ (2013) review already discussed in the section on handedness also discussed several articles on biases in social gesturing. They’ve found 8 additional studies reporting higher incidences of RH for social gesturing than fine motor control in baboons, chimpanzees, and bonobos. No studies that failed to find this difference were reported. Using a continuous measure called the handedness index, the researchers compared studies on fine motor control and studies on social gesturing. The handedness index is a value between -1 (total left-handedness) and 1 (total right-left-handedness) and is based on the total amount of RH responses. The strength of RH displayed on average by baboons, chimpanzees and bonobos was about 2.5 to 3.6 greater for social gestures than for manual tasks. Taken together, there is substantial evidence for a left hemisphere lateralization of social gesturing that exceeds the lateralization of general fine motor
control. This strongly contradicts qualitative change theory, mild version or not.
Data on vocal communication generally supports this conclusion. Heffner & Heffner already showed in 1984 that damage to the left superior temporal lobe of macaque monkeys produces symptoms similar to those that develop in humans after damage to Wernicke’s area. In humans, damage to these areas impairs the ability to understand language significantly. In macaque monkeys, the ability to respond correctly to different species specific calls was found to be impaired. Just like in humans, these impairments were not present when the same area in the right hemisphere was damaged. This data already strongly suggests that vocal communication is left hemisphere
lateralized in primates. Behavioral research adds to this suggestion by finding that the biases found in responses to dichotic listening task in humans also exist in macaque monkeys (Beecher et. al., 1979; Petersen et. al., 1984). Research on neuroanatomy has found some similarities in
neuroanatomy between great apes and humans theorized to underlie our language abilities.
Enlargement of the primate homologue of broca’s area in the left hemisphere compared to the right was present in a sample of 20 chimpanzees, 5 bonobos and 2 gorillas (Cantaloup & Hopkins, 2001). The same asymmetries were found in Wernicke’s area in 12 chimpanzees (Spocter et. al., 2010). The left sylvian fissure is longer than the right in primates as well as humans, and this is theorized to be important for language abilities because the overall size of the temporal lobe (and thus Wernicke’s area) is theorized to be related to length of this fissure (LeMay & Geschwind, 1975). Finally, research using neuroimaging techniques has shown that species specific vocalizations activate areas in the left hemisphere more than in the right. Production of calls activates the primate homologue of Broca’s area in the left hemisphere of chimpanzees (Taglialatela et. al., 2008). Perception of calls activates the left hemisphere more than the right in general (Poremba et. al., 2004), and the primate homologues of both Broca’s and Wernicke’s area specifically in macaque monkeys (Gil-da-Costa et. al., 2006). One neuroimaging study found a right hemisphere bias for perception of calls in
chimpanzees however (Taglialatela et. al, 2009), but taken together the evidence for left hemisphere lateralization of vocal communication outweighs the evidence against (for a summary, see Table 4). The milder version of qualitative change theory is therefore not supported by these results.
The data presented in this paragraph has shown that qualitative change theory is unlikely to be true. First, the presence of some form merge ability and some capability of understanding arbitrariness of symbols contradicts the claim that primates do not possess any language abilities. Second, both gestural and vocal precursors of language used by primates already show left hemisphere lateralization. Therefore, a sudden emergence of language in humans cannot have caused left hemisphere lateralization for fine motor control. The data in this paragraph also sheds some new light on the conclusion drawn about gestural theory in the last section. Gestural theory states that left hemisphere lateralization of general fine motor control in primates caused LHLL in humans later in evolution. No distinction is made between fine motor control used for social
gesturing or general fine motor control. This distinction is present though: left hemisphere lateralization of general fine motor control on the population level is weak and hard to observe, whereas left hemisphere lateralization of social gesturing is much stronger. This implies that either social gesturing has aided the development of general RH, or that the two types of fine motor control are independent. Neither option is supported by gestural theory. Moreover, vocal language abilities of primates also already show lateralization. Since gestural theory proposes a sequence of events in which RH came before LHLL, this finding also contradicts gestural theory. Fine motor control and language lateralization seem independent from one another. This paragraph therefore most strongly supports general mechanism theory.
A much stronger population wide preference for side is found in humans for both lateralization of language and fine motor control that is not observed in primates in any of the mentioned studies. It is not clear whether this stronger lateralization is a consequence of stronger general lateralization of humans. Possibly, lateralization of fine motor control and language have become interconnected in our brains and strengthened each other, even if this connection did not exist yet in our ancestors. If this is the case, language lateralization and lateralization of fine motor control should be related to each other in humans. Furthermore, if the organization of one ability influences the effectivity of organization of the other (as both gestural and qualitative change theory suggest, but general mechanism theory does not), it would be reasonable to expect that atypical organization of one ability is has an effect on performance in the other. These subjects will be discussed in the next paragraph.
§4. Fine Motor Control and Language in Humans
Data on lateralization of fine motor control and language in our ancestors support neither gestural nor qualitative change theory. General mechanism theory is not contradicted by these data, but interactions between language and handedness were not explicitly investigated. Independence of lateralization of language and fine motor control can be investigated in humans in two ways. First, genetic influences associated with language lateralization and handedness will be investigated. Next, behavioral data on language ability and fine motor control can be compared between people with different lateralization patterns.
§4. 1 Genetics and Development of Handedness and Language Lateralization
All theories discussed so far assume that our population wide preference for side of
lateralization is a result of natural selection and therefore heritable. This assumption is not shared by all accounts of handedness and language lateralization however. Provins (1997) for example
suggests that the proportion of people preferring the right hand is actually much smaller than the 90% that is often reported because many left-handed children are successfully forced to become right-handed by the right-biased society they grow up in. LHLL might be a consequence of this too (note that this theory is in line with gestural theory, but proposes a social basis of our population wide RH instead of an inherited one). This would be problematic for all theories discussed so far. Heritability can be examined using several paradigms. The most popular method, twin studies, will not be extensively discussed here. In a recent review, Ooki (2014) discussed that the only conclusion that can be drawn from twin studies performed between 1924 and 2014 is that NRH is more common among twinborn individuals than among singleborn individuals. Estimators of the
contribution of genetic background versus environment to handedness and language lateralization vary widely, most likely because the field suffers from chronic use of small sample sizes and high inconsistency in study designs. We will therefore focus on another methods of estimating heritability: development of the trait.
Studies from the field of developmental psychology show that our left hemisphere preference is already present in infancy, before the development of language or fine motor skills. Nelson, Campbell and Michel (2013) tested handedness using a bimanual task in 34 children at 6 to 14 months and again at 18 to 24 months of age. Based on the proportion of spontaneous use of the right hand for finer movements the children were divided in to RH, LH and AH. The results showed that RH preference was already present in 39% of the infants. The other 61% however was AH. When the children reached toddler age 76% of them showed a RH preference (all infants with RH preference also showed RH preference as toddlers), while 21% was LH and 3% AH. RH preference therefore indeed seems to be the default, but the higher incidence of LH among toddlers and lower incidence of AH suggests that some NRH people do learn to become more RH at later ages. Similar RH/LH ratios (indicating that RH is about 3-4 times more common than LH) and developments were reported for infants and toddlers by Michel, Tyler, Ferre and Sheu (2006), Fagard, Spelke, and von Hofsten (2009) and Esseily, Jacquet & Fagard (2011), though the amount of AH children differed greatly per study. This difference is possibly related to the different methods used by the
researchers, perhaps in similar ways as have been discussed in the section on primate handedness. Hepper, Wells and Lynch (2005) researched laterality in behavior of fetuses. Their data showed that 80% of their 75 fetuses already showed a preference for their right thumb, while 20% showed a preference of the left thumb (AH was not considered an option in their analysis).
Interestingly, all RH fetuses were still RH as 10 -12 year old children but of LH fetuses 33% was classified as RH at that age. These results strongly support the existence of a genetic basis for handedness. Most children already show a strong preference, even prenatally, before the
development of any serious fine motor skills. Lateralization of fine motor skills therefore does seem heritable. Our strong population wide RH preference is already present at a very young age as well, but does seem to increase significantly with age. Curiously, the mentioned studies imply that only LH children switch handedness. This suggests that external pressures to use the right hand might increase the RH/LH ratio, providing some support for social models of handedness. On the other hand, it’s also possible that a child starts to develop RH only when the complexity of motor skills demanded from the child increases. In both cases, it is clear that even without social pressures the majority of infants is RH. Though these studies do not show that the dominance of RH is a stable phenomenon across all age groups, the fact that dominance of RH is also present (even more strongly) in adults strongly suggests that it is. These data therefore show that it’s very probable that handedness (and our population wide RH preference as well) is a heritable trait in humans.
Neuroimaging studies on language lateralization show similar trends as those on
handedness: an increase in lateralization with age but some LHLL consistency in early life (Szaflarski, Holland, Schmithorst, & Byars, 2006; Holland et. al., 2007; Ressel et. al. 2008). The increase of language laterality occurs later than the increase in handedness however and can be observed until age 20 (Szaflarski, Holland, Schmithorst, & Byars, 2006). LHLL also emerges at a different time than handedness. A neuroimaging study on language comprehension performed on 20 babies younger than four days old observed no differences in activation between the left and right hemisphere when listening to familiar and unfamiliar speech sounds (May, Byers-Heinlein, Gervain & Werker, 2011). Shultz, Vouloumanos, Bennett and Pelphrey (2014) on the other hand found a left lateralized
