Learning Repetition, but not Syllable Reversal

Reduplication is common, but analogous reversal processes are rare, even though reversal, which involves nested rather than crossed dependencies, is less complex on the Chomsky hierarchy. We hypothesize that the explanation is that repetitions can be recognized when they match and reactivate a stored trace in short-term memory, but recognizing a reversal requires rearranging the input in working memory before attempting to match it to the stored trace. Repetitions can thus be recognized, and repetition patterns learned, implicitly, whereas reversals require explicit, conscious awareness. To test these hypotheses, participants were trained to recognize either a reduplication or a syllable-reversal pattern, and then asked to state the rule. In two experiments, above-chance classification performance on the Reversal pattern was confined to Correct Staters, whereas above-chance performance on the Reduplication pattern was found with or without correct rule-stating. Final proportion correct was positively correlated with final response time for the Reversal Correct Staters but no other group. These results support the hypothesis that reversal, unlike reduplication, requires conscious, time-consuming computation.

Conclusion

We summarize the argument of the book: the case for taking moral uncertainty seriously, and the case for an information-sensitive account: the correct rule for making decisions under moral uncertainty depends crucially on the information provided by the moral theories in which one has credence. We then note some open questions that would require further work to address, such as how to axiomatize decision-making under moral uncertainty, how we should assign deontic statuses, such as permissible and impermissible, under moral uncertainty, and what a reasonable credence distribution across different moral theories looks like. We end by considering what implications our argument has for the value of doing moral philosophy, suggesting we should conclude that further normative research is one of the most important moral priorities of our time.

Extrapolation accuracy underestimates rule learning: Evidence from the function-learning paradigm

Understanding the development of non-linear processes such as economic or populationgrowth is an important prerequisite for informed decisions in those areas. In the function-learningparadigm, people’s understanding of the function rule that underlies the to-be predicted process istypically measured by means of extrapolation accuracy. Here we argue, however, that even thoughaccurate extrapolation necessitates rule-learning, the reverse does not necessarily hold: Inaccurateextrapolation does not exclude rule-learning. Experiment 1 shows that more than one third of participants who would be classified as “exemplar-based learners” based on their extrapolation accuracy were able to identify the correct function shape and slope in a rule-selection paradigm, demonstrating accurate understanding of the function rule. Experiment 2 shows that higher proportions of rule learning than rule-application in the function learning paradigm is not due to (i) higher a priori probabilities to guess the correct rule in the rule-selection paradigm; nor is it due to (ii) a lack of simultaneous access to all function values in the function-learning paradigm. We conclude that rule application is not tantamount to rule-learning, and that assessing rule-learning via extrapolation accuracy underestimates the proportion of rule learners in function-learning experiments.

Reward associations do not explain transitive inference performance in monkeys

Most accounts of behavior in nonhuman animals assume that they make choices to maximize expected reward value. However, model-free reinforcement learning based on reward associations cannot account for choice behavior in transitive inference paradigms. We manipulated the amount of reward associated with each item of an ordered list, so that maximizing expected reward value was always in conflict with decision rules based on the implicit list order. Under such a schedule, model-free reinforcement algorithms cannot achieve high levels of accuracy, even after extensive training. Monkeys nevertheless learned to make correct rule-based choices. These results show that monkeys’ performance in transitive inference paradigms is not driven solely by expected reward and that appropriate inferences are made despite discordant reward incentives. We show that their choices can be explained by an abstract, model-based representation of list order, and we provide a method for inferring the contents of such representations from observed data.

Meaningful patterns of information in the brain revealed through analysis of errors

SummaryGreat excitement has surrounded our ability to decode task information from human brain activity patterns, reinforcing the dominant view of the brain as an information processor. We tested a fundamental but overlooked assumption: that such decodable information is actually used by the brain to generate cognition and behaviour. Participants performed a challenging stimulus-response task during fMRI. Our novel analyses trained a pattern classifier on data from correct trials, and used it to examine stimulus and rule coding on error trials. There was a striking interaction in which frontoparietal cortex systematically represented incorrect rule but correct stimulus information when participants used the wrong rule, and incorrect stimulus but correct rule information on other types of errors. Visual cortex, by contrast, did not code correct or incorrect information on error. Thus behaviour was tightly linked to coding in frontoparietal cortex and only weakly linked to coding in visual cortex. Human behaviour may indeed result from information-like patterns of activity in the brain, but this relationship is stronger in some brain regions than in others. Testing for information coding on error can help establish which patterns constitute behaviourally-meaningful information.

Introduction

This chapter introduces the distinction between normative internalism and normative externalism, and introduces the kinds of arguments that will appear throughout the book. The main worked example in the chapter concerns uncertainty about the correct rule for decision-making. It is shown that trying to find a rule for decision-making that is sensitive to one’s uncertainty about what the right rule for decision-making is will lead to vicious regresses. The chapter also discusses historical and contemporary views that have anticipated or motivated the views put forward in Normative Externalism, and discusses why the term ‘externalism’ is being applied to yet another philosophical view.

Reward associations do not explain transitive inference performance in monkeys

The observation that monkeys appear to make transitive inferences has been taken as evidence of their ability to form and manipulate mental representations. However, alternative explanations have been proposed arguing that transitive inference performance based on expected or experienced reward value. To test the contribution of reward value to monkeys’ behavior in TI paradigms, we performed two experiments in which we manipulated the amount of reward associated with each item in an ordered list. In these experiments, monkeys were presented with pairs of items drawn from the list, and delivered rewards if subjects selected the item with the earlier list rank. When reward magnitude was biased to favor later list items, correct responding was reduced. However, monkeys eventually learned to make correct rule-based choices despite countervailing incentives. The results demonstrate that monkeys’ performance in TI paradigms is not driven solely by expected reward, but that they are able to make appropriate inferences in the face of discordant reward associations.

Reward associations do not explain transitive inference performance in monkeys

The observation that monkeys appear to make transitive inferences has been taken as evidence of their ability to form and manipulate mental representations. However, alternative explanations have been proposed arguing that transitive inference performance based on expected or experienced reward value. To test the contribution of reward value to monkeys’ behavior in TI paradigms, we performed two experiments in which we manipulated the amount of reward associated with each item in an ordered list. In these experiments, monkeys were presented with pairs of items drawn from the list, and delivered rewards if subjects selected the item with the earlier list rank. When reward magnitude was biased to favor later list items, correct responding was reduced. However, monkeys eventually learned to make correct rule-based choices despite countervailing incentives. The results demonstrate that monkeys’ performance in TI paradigms is not driven solely by expected reward, but that they are able to make appropriate inferences in the face of discordant reward associations.

Reward associations do not explain transitive inference performance in monkeys

The observation that monkeys appear to make transitive inferences has been taken as evidence of their ability to form and manipulate mental representations. However, alternative explanations have been proposed arguing that transitive inference performance based on expected or experienced reward value. To test the contribution of reward value to monkeys’ behavior in TI paradigms, we performed two experiments in which we manipulated the amount of reward associated with each item in an ordered list. In these experiments, monkeys were presented with pairs of items drawn from the list, and delivered rewards if subjects selected the item with the earlier list rank. When reward magnitude was biased to favor later list items, correct responding was reduced. However, monkeys eventually learned to make correct rule-based choices despite countervailing incentives. The results demonstrate that monkeys’ performance in TI paradigms is not driven solely by expected reward, but that they are able to make appropriate inferences in the face of discordant reward associations.

Look again: Pedagogical demonstration facilitates children's use of counterevidence

In learning about the world, we must not only make inferences based on minimal evidence, but must deal with conflicting evidence and question those initial inferences when they appear to be wrong. In three experiments (N=96), we found that in some cases young children only revise their causal beliefs when conflicting evidence is explicitly demonstrated for them. Four- and 5-year-old children inferred a rule about what objects had causal powers, and then saw evidence conflicting with that initial inference. Critically, the conflicting evidence was produced either instrumentally and intentionally, or demonstrated communicatively and pedagogically. Only when evidence was explicitly demonstrated for them did children revise their initial hypothesis and use a subtle clue to infer the correct rule.