Do Babies Think or Know Anything?
Part 1 of the Brain & Mind Series
“The baby, assailed by eyes, ears, nose, skin, and entrails at once, feels it all as one great blooming, buzzing confusion”
William James (1842-1910)
Infants may understand, and think about, the world a lot more than we realise – they just can’t voice it. For example, evidence suggests infants can understand other’s intentions, judge other’s dancing, and know that it’s impossible for an object to disappear out of thin air. Scientists can find out what infants think and understand about the world by measuring how long they look (i.e. their looking time) at implausible events compared to plausible events (see famous example below).
Experiments using this technique are called looking-time studies. This paper considers the limitations of looking-time studies that aim to detect what infants think and know about the world around them. The most serious limitation is that infants may stare at something for various reasons and we can’t truly know why. Another issue is that there is disagreement regarding what tasks should be used and how old infants should be when taking part in these experiments. However, since findings are often non-significant, researchers tend to focus on consistent findings to draw conclusions. This paper discusses those consistencies. It is then argued that infants have a basic understanding of physical principles, such as the principle that an object cannot disappear out of thin air. Infants may also possess a basic understanding of the causes of other’s behaviour (i.e. intentions). Infants’ basic understanding is then enhanced through their later experiences. This paper therefore demonstrates what looking-time studies can, and cannot, tell us about infants’ thoughts and understanding.
Here’s my informed opinion – what’s yours?
I think that it is easy to project our own thoughts and feelings on to infants and therefore incorrectly infer that they understand the world around them. As such, I think some researchers might be reading too much in to a look. For example, quite often I stare in to space and might look as if I am thinking deeply, but there might not be much going on inside my head. On the other hand, I also think that in the past psychologists have under-estimated infants’ abilities. I am convinced by the research demonstrating that infants have a basic understanding of the physical and social world. Overall, I do think looking-time studies can reveal the inner workings of infants’ minds to an extent, but I only get excited about findings that are consistent.
Feel free share your opinion in the comments section below.
What conclusions about infant ‘thought’ and ‘understanding’ can we draw from looking time data?
The field of developmental psychology has been heavily reliant upon looking-time data, since it is easily obtainable from preverbal infants due to its low motor demands. Over the last few decades, vast amounts of looking-time data have been collected which have informed theories of infant cognition. This essay will consider what conclusions can be drawn from looking-time data regarding infants’ thoughts and understanding of the social and physical world. This will be achieved through an overview of the limitations of looking-time studies, as well as a discussion of whether these prevent findings from being conclusive. It will also consider the regularities in looking-time data that may allow for accurate conclusions to be derived.
There exists a diverse array of claims drawn from looking-time data regarding infants’ thought processes and understanding. Recent claims include that infants can reason about the social nature of food (Liberman, Woodward, Sullivan & Kinzler, 2016), understand others’ needs (Köster, Ohmer, Nguyen & Kärtner, 2016), and identify bad dancing (Hannon, Schachner & Nave-Blodgett, 2017). From a surface-level perspective, one might deduce that looking-time is a versatile measure of infants’ cognitive abilities, or rather that researchers are over-estimating its informative power. The collective view in the field is that looking-time studies have been informative, but possess weaknesses (Aslin, 2007; Cantrell & Smith, 2013) constraining the conclusions that can be drawn.
The most fundamental of these limitations resides in the underlying assumption of all looking-time paradigms (e.g. preferential looking, habituation-dishabituation), which is that they can reveal cognitive abilities. Arguably, this assumption exceeds the scope of evidence due to the ambiguous nature of looking-time as a dependent variable. For instance, increased looking-time may reflect a range of cognitive processes including sustained attention with positive, null or negative affect (Aslin, 2007). Additionally, the assumption is potentially unfalsifiable, because we cannot directly access infants’ reasoning and knowledge via language, hence why looking-time measures are used. Yet, theories concerning infants’ understanding of the physical and social world have been driven by looking-time data; predominately obtained using the violation-of-expectation paradigm. This paradigm measures looking-time in response to expected/plausible events versus an unexpected/implausible event designed to violate a physical/psychological principle or an experimentally-induced expectation. It is assumed that if an infant thinks an event/display is unusual, they will look at it for a prolonged duration because it is incongruent with their implicit understanding of the world (Baillargeon, Scott & Bian, 2016).
Furthermore, with all such paradigms there is a lack of consensus regarding how many trials should be averaged to enable statistical comparisons, and the inclusion criteria for the final sample (Cantrell & Smith, 2013). Because of these issues, and high attrition rates, findings obtained from looking-time studies may lack generalisability. It is also argued that there is a high degree of variability in observed effects, and the parameters inducing them may change or be a function of the tasks/stimuli used, or the age of the sample (Cantrell & Smith, 2013). Collectively, these issues jeopardise the validity of conclusions drawn from looking-time data.
However, Cantrell and Smith (2013) have argued that since it is easy to obtain null results using looking-time data, any statistically reliable effect, or replicated effects, may be meaningful. Due to this, the field generally focuses on consistent findings rather than deliberating over the methodological features of looking-time studies (Cantrell & Smith, 2013). Such consistencies may allow us to derive conclusions regarding the extent of infants’ understanding of the social and physical world. One of the most consistent findings is that by 12 months, infants seemingly possess an understanding that intentional, or goal-directed, mental states underpin other’s actions and expressions (Dunphy-Lelii, LaBounty, Lane & Wellman, 2014). This conclusion is derived from a host of looking-time findings indicating that infants can ascribe intentions, perceptions and beliefs to agents, permitting the infant to understand the behaviour of others (see review by Luo & Baillargeon, 2010).
In line with this, an extensive review of looking-time studies by Baillargeon et al. (2016) concluded that infants are able to infer an agent’s motivational, epistemic, and counterfactual states. It has been further concluded that infants are able to use these mental states to interpret, and make predictions, concerning an agent’s future actions (Baillargeon et al., 2016). However, while these claims appear conclusive due to the extent of support from looking-time studies, many other claims cannot be deemed conclusive, such as those initially described, as they require further confirmatory evidence. This is often the case within psychological sciences more generally.
A further common challenge within the psychological sciences, is that of conflicting evidence. This issue afflicts looking-time studies investigating infants’ understanding of the physical world. For instance, evidence has indicated that infants as young as 4 months understand that objects are separate entities, and the factors that constrain object motion (Spelke, Breinlinger, Macomber & Jacobson, 1992; Spelke, Kestenbaum, Simons & Wein, 1995), but this understanding can only be demonstrated under certain circumstances (Baillargeon, 2002; Schilling, 2000; Wang, Baillargeon & Paterson, 2005). These constraining factors include solidity, which refers to the way in which two objects cannot exist in the same location, and continuity, which refers to the way in which objects cannot randomly appear or disappear (Baillargeon et al., 2012).
To explain these discrepancies, a new theoretical account has been proposed, contending that an understanding of physical principles is developed via experience across infancy. It holds that infants possess an innate physical-reasoning system which provides the necessary causal framework to facilitate this development (Baillargeon et al., 2012). This framework includes several core principles, including those aforementioned, that are thought to be elaborated on during the acquisition of causal rules which allow for the outcomes of physical events to be predicted and interpreted (Baillargeon et al., 2012). This theory also accounts for findings indicating that the conceptual understanding of inertia and gravity, viewed as extensions of the principle of continuity, develop later in infancy (Kim & Spelke, 1992; Spelke, Katz, Purcell, Ehrlich & Breinlinger, 1994).
Spelke discussing how infants build on their basic understanding of the world
Collectively, these findings and theories drawn from looking-time data have promoted ever-increasing confidence in infants’ cognitive abilities. However, in light of some of the claims initially presented, it could be argued that researchers are now over-estimating infants’ abilities, particularly given that the fundamental issue of looking-time data being ambiguous in nature still holds. For instance, as alluded to previously, there is an ongoing debate concerning whether lower-level explanations suffice that do not necessitate infants possessing a precursory understanding. An example of this is the pre-attentive model which poses that infants’ looking-times vary solely as a function of visual processing at the pre-attentive level (Lovett & Scassellati, 2004).
However, the field of robotics may be able to resolve this. This is since, if robots, that do not possess an understanding of higher-level concepts, can replicate infant behaviour, it would suggest that theorists are over-estimating infants’ understanding (Lovett & Scassellati, 2004). This in turn would allow for claims to be tested. For example, Lovett and Scassellati (2004) tested the validity of the pre-attentive model by conducting Baillargeon’s (1986) experiment, involving the violation-of-expectation paradigm, on a humanoid robot. In the original experiment, and others that have followed such as that by Wang, Baillargeon and Brueckner (2004), infants appeared to understand object permanence which denotes the principle that objects do not cease to exist when hidden. Lovett and Scassellati (2004) found comparable results, thus indicating that Baillargeon’s (1986) original findings can instead be explained using the pre-attentive model. Whilst this model still requires further validation, this finding does indicate that theorists are over-estimating infants’ understanding of the physical world.
Nevertheless, such research is in its infancy and there are numerous other low level explanations for looking-time findings that could be tested. Moreover, as is the case across all the psychological sciences, deriving conclusions from only one type of data can constrain both the number and validity of those conclusions. For instance, it is as yet unknown how infants are able to infer and reason about others’ mental states, how this ability develops, nor how the brain networks underpinning this development mature (Baillargeon et al., 2016). So, while looking-time data may suggest the presence of infants’ implicit understanding and thought processes, it cannot inform us of the underlying mechanisms by which they emerge and operate.
Furthermore, it is only possible to verify the accuracy of conclusions from looking-time data by contrasting the findings with those obtained using alternative, or convergent methods, which are not reliant upon the same assumptions. For example, research carried out by Dunphy et al. (2014) and Hood, Carey and Prasada (2000) have validated and challenged the conclusions drawn respectively. Dunphy et al. (2014) examined whether looking-time measures accurately depict infants’ understanding of social relationships. It was found that infants’ performance on a looking-time task, designed to measure their understanding of false belief, was related to how well they socially engaged with their mothers. Hood et al. (2000) found, in line with many observational studies using search tasks, that two-year-olds are unable to demonstrate their supposed understanding of physical principles, such as solidity. Moreover, with the use of more advanced eye-tracking devices, neuroimaging methods, and robotics, we will be better equipped to explain such discrepancies, as well as test and supplement the conclusions drawn from looking-time data.
In conclusion, looking-time data has revealed that infants appear to think about, and understand, fundamental aspects of the social and physical world on an implicit level. Such findings include those indicating that infants have a rudimentary understanding of the underlying causes of others’ behaviour, as well as an innate ability to reason about basic physical principles which may allow for an understanding of more complex principles to later develop. Therefore, conclusions can be drawn from looking-time data, despite its inherent constraints, but may lack validity if drawn from this data alone.
Interested in your baby taking part in experiments? Visit UCL’s baby lab website
Aslin, R. N. (2007). What’s in a look?. Developmental science, 10(1), 48-53.
Baillargeon, R. (1986). Representing the existence and the location of hidden objects: Object permanence in 6-and 8-month-old infants. Cognition, 23(1), 21-41.
Baillargeon, R. (2002). The Acquisition of Physical Knowledge in Infancy: A Summary in Eight Lessons. In U. Goswami (Ed.), Blackwell Handbook of Childhood Cognitive Development (pp. 47-83). Oxford: Blackwell.
Baillargeon, R., Scott, R. M., & Bian, L. (2016). Psychological reasoning in infancy. Annual Review of Psychology, 67, 159-186.
Baillargeon, R., Stavans, M., Wu, D., Gertner, Y., Setoh, P., Kittredge, A. K., & Bernard, A. (2012). Object individuation and physical reasoning in infancy: An integrative account. Language Learning and Development, 8(1), 4-46.
Cantrell, L., & Smith, L. B. (2013). Open questions and a proposal: A critical review of the evidence on infant numerical abilities. Cognition, 128(3), 331-352.
Dunphy-Lelii, S., LaBounty, J., Lane, J. D., & Wellman, H. M. (2014). The social context of infant intention understanding. Journal of Cognition and Development, 15(1), 60-77.
Hannon, E. E., Schachner, A., & Nave-Blodgett, J. E. (2017). Babies know bad dancing when they see it: Older but not younger infants discriminate between synchronous and asynchronous audiovisual musical displays. Journal of Experimental Child Psychology, 159, 159-174.
Hood, B., Cole-Davies, V., & Dias, M. (2003). Looking and search measures of object knowledge in preschool children. Developmental psychology, 39(1), 61.
Kim, I. K., & Spelke, E. S. (1992). Infants’ sensitivity to effects of gravity on visible object motion. Journal of Experimental Psychology: Human Perception and Performance, 18(2), 385-393.
Köster, M., Ohmer, X., Nguyen, T. D., & Kärtner, J. (2016). Infants understand others’ needs. Psychological science, 27(4), 542-548.
Liberman, Z., Woodward, A. L., Sullivan, K. R., & Kinzler, K. D. (2016). Early emerging system for reasoning about the social nature of food. Proceedings of the National Academy of Sciences, 113(34), 9480-9485.
Lovett, A., & Scassellati, B. (2004). Using a robot to reexamine looking time experiments. In Proceedings of the Third International Conference on Development and Learning (pp. 284-291).
Luo, Y., & Baillargeon, R. (2010). Toward a mentalistic account of early psychological reasoning. Current Directions in Psychological Science, 19(5), 301-307.
Schilling, T. H. (2000). Infants’ looking at possible and impossible screen rotations: The role of familiarization. Infancy, 1(4), 389-402.
Skerry, A. E., & Spelke, E. S. (2014). Preverbal infants identify emotional reactions that are incongruent with goal outcomes. Cognition, 130(2), 204–216. http://doi.org/10.1016/j.cognition.2013.11.002
Spelke, E. S., Breinlinger, K., Macomber, J., & Jacobson, K. (1992). Origins of knowledge. Psychological review, 99(4), 605-632.
Spelke, E. S., Katz, G., Purcell, S. E., Ehrlich, S. M., & Breinlinger, K. (1994). Early knowledge of object motion: Continuity and inertia. Cognition, 51(2), 131-176.
Spelke, E.S., Kestenbaum, R., Simons, D., & Wein, D. (1995). Spatiotemporal continuity, smoothness of motion and object identity in infancy. The British Journal of Developmental Psychology, 13, 113–142.
Wang, S. H., Baillargeon, R., & Brueckner, L. (2004). Young infants’ reasoning about hidden objects: evidence from violation-of-expectation tasks with test trials only. Cognition, 93(3), 167-198.
Wang, S. H., Baillargeon, R., & Paterson, S. (2005). Detecting continuity violations in infancy: A new account and new evidence from covering and tube events. Cognition, 95(2), 129-173.