Understanding how infants perceive the world around them is an important aspect of understanding human cognition. Recent studies have delved deep into this realm, discovering that even before they can speak, infants have a grasp on abstract relations like "same" and "different."
Abstract relations stand at the pinnacle of human thought, enabling us to engage in analogical and logical reasoning. This capability, which might differentiate humans from other species, is present even in infants. Through a series of experiments involving pupillometry—a technique measuring pupil dilation in response to cognitive load—researchers have shown that infants, specifically those aged 10 to 12 months, can understand the concept of "same."
Interestingly, the infants' ability to comprehend the relation of the "same" is not without its limits. It appears closely tied to their working memory capacity, which at this age can handle about three to four items. Infants successfully recognize when four syllables are the same and can even generalize this to new sequences. However, their performance dips when the sequence extends to five or six syllables, highlighting a crucial constraint in their cognitive processing.
One might wonder how infants represent this abstract relation in their minds. The current consensus leans away from the idea of discrete symbols, at least for pre-lexical infants. Instead, their understanding of "same" seems to be constructed by piecing together symbols for individual entities. This form of representation, while abstract enough to allow generalization across different entities, is distinct from adult-like propositional representations.
The journey from recognizing abstract relations to applying them in tasks presents its hurdles. For instance, young children and even non-human animals struggle with relational match-to-sample tasks (RMTS) (see https://www.lesswrong.com/posts/et3SCxGqXDfJdeP5f/relational-thinking-in-animals-and-humans), which require matching based on abstract relations rather than perceptual similarities. This difficulty suggests a developmental gap in representing the "same" in a discrete, symbol-independent manner.
However, the capacity of an infant's working memory plays a pivotal role in their ability to grasp abstract relations. This study's findings indicate that exceeding this capacity—by introducing sequences with more than four syllables—impairs their performance. Such results underscore the intertwined nature of working memory limitations and abstract relation processing in infancy.
Empirical Findings
Recent research has taken us a step closer to understanding how infants, those between 10 to 12 months old, process abstract concepts like "same."
With 192 infants, researchers divided them into groups across six experiments. These infants were exposed to sequences of syllables, with the tasks designed to test their ability to detect and generalize the relation of "same." They used 48 unique syllables generated by a speech synthesizer and presented them in sequences while recording infants' reactions using an eye tracker.
Experiment 1
In the first experiment, infants were given sequences of four syllables. The aim was to see if they could recognize when all syllables were the same (AAAA) versus when one was different (AAAB). Results showed infants noticed the difference in sequences with one odd syllable, indicating they can grasp the "same" concept with up to four items.
Experiment 2
The second experiment increased the sequence to five syllables to test the limits of infants' working memory. The sequences were either all the same (AAAAA) or had one different syllable (AAAAB). In this setup, infants did not show a significant reaction to the different syllables, suggesting a challenge in processing more than four items as "same."
Experiment 3
Experiment 3 further pushed the limit to six syllables. Like before, sequences were either all the same (AAAAAA) or slightly different (AAAAAB). The infants reacted to the different syllables, but this seemed driven by a simple change in the sequence rather than an understanding of the "same" relation.
Experiment 4
In the fourth experiment, researchers modified where the different syllables appeared in a six-syllable sequence (AAABAA instead of AAAAAA or AAAAAB). Here, infants did notice the different syllables, suggesting they could track the "same" relation within the first four syllables of a longer sequence.
Experiment 5 & 6
The final two experiments tested whether infants could generalize the concept of "all same" across sequences of varying lengths (three, four, and five syllables in Experiment 5 and two to four syllables in Experiment 6). The infants did not show recognition of the "all same" concept across different lengths, indicating difficulty in abstracting the "same" relation beyond specific instances.
Infants can recognize the "same" relation in sequences of up to four syllables. Their ability to process this abstract relation seems capped by the limits of their working memory. (Oddly, this is very reminiscent of the second core system of numbers which only represents up to 4 https://www.lesswrong.com/s/9YkW5gc8pio55CGy6/p/kfoqLxnWdLrwsPGG8)
More than four syllables present a challenge. When sequences extended beyond four syllables, infants struggled to apply the "same" concept, likely due to working memory constraints.
Infants need consistency in sequence length. Changing the length of the "same" sequences confused the infants, suggesting they rely on consistent patterns to understand abstract relations.
Understanding how infants perceive the world around them is an important aspect of understanding human cognition. Recent studies have delved deep into this realm, discovering that even before they can speak, infants have a grasp on abstract relations like "same" and "different."
Abstract relations stand at the pinnacle of human thought, enabling us to engage in analogical and logical reasoning. This capability, which might differentiate humans from other species, is present even in infants. Through a series of experiments involving pupillometry—a technique measuring pupil dilation in response to cognitive load—researchers have shown that infants, specifically those aged 10 to 12 months, can understand the concept of "same."
Interestingly, the infants' ability to comprehend the relation of the "same" is not without its limits. It appears closely tied to their working memory capacity, which at this age can handle about three to four items. Infants successfully recognize when four syllables are the same and can even generalize this to new sequences. However, their performance dips when the sequence extends to five or six syllables, highlighting a crucial constraint in their cognitive processing.
One might wonder how infants represent this abstract relation in their minds. The current consensus leans away from the idea of discrete symbols, at least for pre-lexical infants. Instead, their understanding of "same" seems to be constructed by piecing together symbols for individual entities. This form of representation, while abstract enough to allow generalization across different entities, is distinct from adult-like propositional representations.
The journey from recognizing abstract relations to applying them in tasks presents its hurdles. For instance, young children and even non-human animals struggle with relational match-to-sample tasks (RMTS) (see https://www.lesswrong.com/posts/et3SCxGqXDfJdeP5f/relational-thinking-in-animals-and-humans), which require matching based on abstract relations rather than perceptual similarities. This difficulty suggests a developmental gap in representing the "same" in a discrete, symbol-independent manner.
However, the capacity of an infant's working memory plays a pivotal role in their ability to grasp abstract relations. This study's findings indicate that exceeding this capacity—by introducing sequences with more than four syllables—impairs their performance. Such results underscore the intertwined nature of working memory limitations and abstract relation processing in infancy.
Empirical Findings
Recent research has taken us a step closer to understanding how infants, those between 10 to 12 months old, process abstract concepts like "same."
With 192 infants, researchers divided them into groups across six experiments. These infants were exposed to sequences of syllables, with the tasks designed to test their ability to detect and generalize the relation of "same." They used 48 unique syllables generated by a speech synthesizer and presented them in sequences while recording infants' reactions using an eye tracker.
Experiment 1
In the first experiment, infants were given sequences of four syllables. The aim was to see if they could recognize when all syllables were the same (AAAA) versus when one was different (AAAB). Results showed infants noticed the difference in sequences with one odd syllable, indicating they can grasp the "same" concept with up to four items.
Experiment 2
The second experiment increased the sequence to five syllables to test the limits of infants' working memory. The sequences were either all the same (AAAAA) or had one different syllable (AAAAB). In this setup, infants did not show a significant reaction to the different syllables, suggesting a challenge in processing more than four items as "same."
Experiment 3
Experiment 3 further pushed the limit to six syllables. Like before, sequences were either all the same (AAAAAA) or slightly different (AAAAAB). The infants reacted to the different syllables, but this seemed driven by a simple change in the sequence rather than an understanding of the "same" relation.
Experiment 4
In the fourth experiment, researchers modified where the different syllables appeared in a six-syllable sequence (AAABAA instead of AAAAAA or AAAAAB). Here, infants did notice the different syllables, suggesting they could track the "same" relation within the first four syllables of a longer sequence.
Experiment 5 & 6
The final two experiments tested whether infants could generalize the concept of "all same" across sequences of varying lengths (three, four, and five syllables in Experiment 5 and two to four syllables in Experiment 6). The infants did not show recognition of the "all same" concept across different lengths, indicating difficulty in abstracting the "same" relation beyond specific instances.