Jump to ContentJump to Main Navigation
Causal LearningPsychology, Philosophy, and Computation$

Alison Gopnik and Laura Schulz

Print publication date: 2007

Print ISBN-13: 9780195176803

Published to Oxford Scholarship Online: April 2010

DOI: 10.1093/acprof:oso/9780195176803.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 24 June 2018

(p.353) Index

(p.353) Index

Source:
Causal Learning
Publisher:
Oxford University Press
accidental properties, 190
adjacency consistent-adjacency inconsistent inclusion, 134–36
adjacency consistent error, 134–35
adjacency-integrity, 134–35
agent causal view, 32, 45, 46
alternative causes, unobserved, 284–87
alternatives, contrasting, 106
ambiguous stimuli, dynamic interpretations of, 288–92
Anderson’s rational analysis model, 183
associationist theories, 96. See also associative theories
associative challenge, 141–44
associative learning account of causal learning, 10
associative learning models, 281
associative theories, 94. See also associationist theories
assumptions, 107–8
asymmetry
and control, 250–51
and prediction, 247–50
auditory observation, 45
backtracking vs. non-backtracking counterfactuals, 21, 29
backward blocking, 11, 142–44
Bayes net approaches, 37. See also causal Bayes nets
Bayes net assumption connecting causation and probabilities, 23
Bayes net modeling, 275. See also causal Bayes nets
Bayes net representations, 19–20, 71, 79
Bayes net theory(ies), 75, 99. See also causal Bayes nets
Bayes nets learning algorithm, 72, 82
Bayes’ rule, 315–18
Bayesian framework for causal inference, hierarchical, 315, 320, 340
basic Bayes, 315–16
inferring causal networks, 317–18
inferring causal principles, 318–19
inferring causes and predicting effects, 316–17
Bayesian inference, 143, 216, 315–16, 340
Bayesian methods, 169. See also causal Bayes nets
belief-desire psychology, 261–62
belief-updating model, 283
beliefs
false, 268–70
weighing new evidence against old, 79–80
biologically caused actions, 267
blicket detector, 140, 143, 144, 184, 213, 220, 334–41
graph structures generated by causal theory for, 337, 338
one-cause, 226
probabilistic logical theory for causal induction with, 336, 337
two-cause, 226–27
blocking, 141–42
backward, 11, 142–44
boundary intensification, 195–97
categorical vs. projective inferences, 253
categories
core vs. perceptual, 191
nonessentialized, and theoretical coherence, 200–202
categorization. See also classification
human, 173–74, 186–87. See also similarity functions
speculations about, 182–86
(p.354)
projection and, 252–55
categorization field’s mind-body problem, 191–93, 203
categorization theories, 174–77. See also categorization, models of
category labels, 204–5
category structure
multilayer, 186
one-layer, 185
causal Bayes net formalism, 70, 74–78, 82–83, 169
causal Bayes nets (CBNs), 3–4, 94, 95, 97, 119, 263. See also Bayes net approaches
causal models and, 87–92, 169
causal structure and conditional probabilities, 4–5
deriving predictions from, 87–92
illustrations/examples, 71, 78, 122
interventions and, 5–6
learning, 24, 123–24
learning and, 6–8
normative theory of, 119–20, 136
representational theory of, 129
representing causal systems, 121–23
causal Bayesian networks, 155, 157, 177, 178, 184, 192, 302, 317–18. See also Bayes net approaches
causal chain, 74
causal-chain model, 88, 90, 93–95
causal conditional, 93
causal direction, 105
causal effect, 108. See also total effect
causal grammar(s), 82, 323, 342–44. See also grammar(s); graph grammars
in hierarchical Bayesian framework, 324–25
intuitive theories as, 307–15
toward, 312–14
causal induction, primacy vs. recency effects in, 281–84
causal induction models, 280–81, 288
causal knowledge. See also specific topics
development, 262–63
domain-specific, 9–10
how statistical regularity can translate to, 140–41
causal laws, 337
causal maps, 274–75
causal Markov assumption (CMA), 4–5, 23, 177–79, 235, 240
causal Markov condition (CM), 4–5, 23. See also Markov condition
causal mechanisms, 9–10, 68, 192. See also generative transmission models
causal model theory (CMT), 156, 176, 177, 180, 193
causal modeling, 87–92, 99. See also specific models
integrating fragments from, 166–69
causal models, evolving set of, 205
causal networks, 324, 325, 328. See also causal Bayesian networks
intuitive theories as, 303–5
causal pathways, 107–8
causal power, 330
causal predicates, 336
causal reasoning, 9, 263
vs. logical reasoning, 92–94
normative theory of, 121–24, 136–37
Piagetian account of, 8, 262, 263
vs. propositional logic, 92–94
causal relations, 103–4. See also specific topics
reasons for representing, 245–46, 259
role of interventions in infants’ developing sensitivity to, 50–55
role of interventions in infants’ understanding of physical, 53–55. See also causal understanding
role of interventions in infants’ understanding of psychological, 51–53. See also causal understanding
causal representation, 259
asymmetry, 246–51
underlying mechanism, 246, 251–52, 259
inferential role of information about mechanisms, 252–55
virtues of the mechanism schema, 255–59
causal strength. See strength
causal structure(s), 73
cues to, 157
intervention, 161–63. See also intervention(s)
prior knowledge, 163–66
statistical covariation, 157–59
temporal order, 159–65
learning a wide range of, 73–76
statistical structure and, 249
causal systems, representing. See causal Bayes nets
causal taxonomy, 108–9
causal understanding. See also under causal relations
components, 48
levels, 45–46
mechanisms underlying the development of, 50
causal variables, strength of, 73. See also strength
Causality Lab, 125–27
causal discovery in, 127–28
causation, 101, 294
by reasons, 61–63
cause(s)
multiple, 264
notions of, 110, 111, 248–49
chain graphs, 181
chain model, 158, 159, 162, 166–68
Cheng, Patricia, 298–300
childhood, emotional importance and frequency of explanations in, 265–68
children, determinism in causal inferences of, 213–22
choices
complex, 98–99
simple, 98
Chomsky, Noam, 45, 307–8
class graph, 326
classical conditioning, 26
classical learning, 10–12
classification. See also categorization; essentialism; generative theory of classification
as diagnosis, 193–95
as prospective vs. diagnostic reasoning, 197–200
clique potentials, 179
cognitive cuing explanations, 267
(p.355) cognitive maps, 274–75
common-cause model, 88, 90, 91, 94–96, 98, 99, 165, 168, 209
common-effect model, 88, 90, 158, 165, 167, 168
component effect, 108, 111
computational models of learning, 169–70
conditional independence, 4, 23, 157, 158
conditional intervention principle, 75
conditional probabilities, 4–5, 23
conditioning, 26, 94
instrumental, 26–27, 94
confounded interventions, 80–82
connectionist account of causal learning, 10
connectionist models, 176–77
constraint-based approach, 123, 169, 170
context-specific independence (CSI), 184–86
contingency, 11. See also instrumental learning
contrasting alternatives, 106
control variables, 60–61
counterfactual reasoning, 87
counterfactual theories, 19–20
counterfactuals, 25, 87, 99
backtracking vs. non-backtracking, 21, 29
modeling, 91
covariation, statistical, 157–59, 293. See also alternative causes; ambiguous stimuli; causal induction
covariation account of causal learning, 10
cuing explanations, cognitive, 267
decision making, causal, 97–99
decision theory, 111
dependence, 24, 210. See also independence
conditional vs. unconditional, 4
depression, stressful life events and, 64–65
determinism, 222–30
in children’s causal inferences, 213–24, 334–35
faithfulness, causal inference, and, 208–13
deterministic detector theory, 336–43
deterministic systems, learning the structure of, 231–40
diagnostic learning, 164–65
direct cause (DC), 22–23, 35
direct effects, 108, 111
directed acyclic graphs (DAGs), 249, 250
discovery, logic of, 3
domain-specific causal knowledge, 9–10
dominance (decision-making principle), 98
double effect, doctrine of, 111
effect, law of, 10
efficiency, 255–57
egocentric causal view, 45, 46, 71
egocentric conception, 32–33
egocentrism, infantile, 38–39
equivalence classes. See Markov equivalence classes
error rate, 336
errors, commission vs. omission, 133–34
essences, real vs. nominal, 190
essential properties, 190
evidence
vs. observations and interventions, 77–79
weighing old beliefs against new, 79–80
exemplar-based category structures/models, 175–82
expected utility, maximizing, 98
expected utility theory, 110–11
experimental design, theory of optimal, 128
experimental setup, 123, 126, 129. See also Causality Lab
informative vs. uninformative, 127–28
explanations
of actions, 266–67
in childhood, emotional importance and frequency of, 265–68
foster knowledge acquisition in other domains, 271
of humans vs. objects, 266
implications for theory development and causal learning, 274–76
infants and, 271–73
infant vs. toddler findings, 273
lead development, 268–69
individual differences, 269–70
microgenetic research, 270–71
of mental states, 267–68
vs. predictions, 263–65, 268
faithfulness assumption, 7, 235
false beliefs, 268–70
Fisher, Ronald, 121
focal event, 248
framework theories, 305–7, 323
generalized context model (GCM), 175–79, 182
generative grammar. See grammar(s)
generative theory of classification, 191–93, 202–4. See also essentialism
empirical support for, 193–204
generative transmission, 9–10, 31–32
generative transmission models, 68–69
generative transmission rule, 257
Goldvarg, E., 295
Gopnik, Alison, 263
grammar(s), 320. See also causal grammar(s)
bird’s eye view of generative, 308–12
theories as logical, 333–42
universal, 310, 311, 314
graph grammars. See also causal grammar(s)
theories as, 325–33
graph schemas, 324, 326–27, 352 nn.1,2
examples of, in different domains, 327–29
extensions and limitations, 332–33
learning, 331–32
role of learning causal structure, 329–31
graph surgery, 6, 90
graphical clique, 179
graphical models, 173. See also categorization
applying the graphical model equivalences, 180–82
causal, 3, 304. See also causal Bayes nets
(p.356) habituation paradigm, 52–54, 272
Humean rule, 256–57
Humeanism, 259
identification procedures (categories), 191
imitation and imitative learning, 10, 34–35, 45. See also infant imitation
independence. See also dependence
conditional vs. unconditional, 4, 23
independence relations, 126, 127
infant imitation, 38, 46. See also imitation and imitative learning
innate mapping between observation and execution, 39
goal-directedness, 39–40
innate representation of human action, 40
theories, 38
infants. See also under causal relations; explanations
infer interventions for agents, 42
inferring an intervention based on unsuccessful action patterns, 41–42
agents and goals, 42
inventing new means to achieve an inferred intervention, 42
learning actions vs. outcomes, 41
learning interventions from observation, 40
privileged role for manipulations performed by self, 40–41
learning to use a tool, 43–45
primacy of people in their notion of interventions
agent transformations, 43
involvement of people results in different interpretations of same scene, 42–43
understanding of interventions by self and other, 37–39
understanding of physical and psychological causation, 49–50
instrumental conditioning, 26–27, 94
instrumental learning, 10
intended actions, 267
intervention assumption, 6
interventional reasoning, 87
interventionism, 19–23, 31, 58–60
features, 20–25
interventionist account of causation, 70, 104
interventionist counterfactuals, causal judgment and, 27–29, 347 n.4
intervention(s), 21, 37, 161–63, 227. See also under causal relations; infants; observation(s)
aids learning, 162
Bayes nets and, 5–6
distinguishing evidence from, 77–79
distinguishing good from confounded, 80–82
learning direct causes from, 228–29
learning indirect causes from, 228
making novel, 70–73
modeling, 89–91
social effects, invariance, and, 294–300
vs. temporal order, 163
types of, 63–64, 89
voluntary actions and, 29–30
intuitive theories, 301–4, 319–20, 323, 343–44. See also Bayesian framework for causal inference
as causal grammars, 307–15
as causal networks, 303–5
framework theories and specific theories, 305–7
as hypothesis space generators, 303
three questions about, 301–2
“invisible imitation,” 39
Johnson-Laird, P. N., 295
knowledge-driven approach to causal structure, 317–18
Lagnado, Dave, 294–96
language comprehension, 309
Laplace, Pierre Simon, 208–25
learning order and causal order, 163–64
linguistic grammars, 323. See also grammar(s)
logical conditional, 93
logical theories, learning, 341–42
logical vs. causal reasoning, 92–94
magic, 43
magnetism theory, 328
manipulability, 250–51
manipulated graph, 90, 91, 122, 123
manipulated systems, 122, 123
manipulations, 20–22, 40–41. See also interventionism
Markov assumption. See causal Markov assumption
Markov condition, 107–8, 167. See also causal Markov condition
strong causal, 212
weak causal, 209, 210, 217
Markov equivalence, 157–58
Markov equivalence classes (MECs), 123, 124, 132–36
MEC integrity, 132–34
Markov random field, 179–82
means-ends relationships, 34
Meno, 208–25
mental model theory, contemporary, 295
metacognition, 80–81
mind, theory of, 261–62, 265, 269–70, 274, 276. See also theory theory
mind-body problem, 191–93, 203
mistaken actions, 267
“mixed causes,” 107
modeling, causal. See causal models
modular view of causal reasoning, 9
modus tollens, 93
moral deliberation, 111
morality, 110
morals
for philosophy, 109–11
for psychology, 111–13
multimodal integration, problem of, 149
nativist view of causal reasoning, 9
net effect, 108. See also total effect
net-promoting-cause-on-average, 109–10
newborn imitation. See infant imitation
Newcomb’s paradox, 98
node classes, 326
nominal essences, 190
normative and descriptive theories, 26
novel interventions, making, 70–73
(p.357) observational reasoning, 87
observation(s), 39–41. See also unobservables
auditory, 45
distinguishing evidence from, 77–79
vs. intervention, 21, 39, 86–87, 91, 96, 99, 162
in counterfactual scenarios, 92
modeling, 88–89
predicting outcomes of hypothetical interventions from, 94–96
one-cause trials, 140–41
ontological problem, 121
ontology, 336, 337
operant learning, 10
parameterization, 4
parameterized causal models, reasoning with, 94–96
sensitivity to parameters, 96–97
parent events, 248
parsing, 309
“path diagrams,” 121
path-specific effect, 108
PC algorithm, 233–38
PCD algorithm, 238–40
philosophy and psychology, 25–26
physical causation, infants’ understanding of, 49–50
physically caused actions, 267
Piaget, Jean, 68
Piagetian account of causal reasoning, 8, 262, 263
Piagetian theory of infant imitation, 38–39
posterior probability distribution, 315–16, 330, 331
power theory of probabilistic contrast (PC), 12–13
PowerPC model of causation, 105, 112, 156
practical intelligence, 38
“precausal” reasoning, 8, 262
predicates, causal and structural, 336
prediction. See also under Bayesian framework for causal inference; causal Bayes nets; explanations
asymmetry and, 247–50
vs. explanations, 263–65, 268
manipulability and, 250
predictive learning, 164, 165
“prevent,” 107
primacy vs. recency effects in causal induction, 281–84
primate causal cognition, 30–35
prior knowledge, 163–64
parameter estimation and, 165–66
vs. temporal order, 164–65
use of, 166
prior probability distribution, 315
priors. See setting priors
probabilistic contrast (PC), 12–13
probabilistic detector theory, 338, 340–43
probability distribution
posterior, 315–16, 330, 331
prior, 315
processing constraints, 166
projection, 252–55
promoting vs. preventing/inhibiting causes, 105
propositional logic vs. causal reasoning, 92–94
prototype-based categorization models, 175–82, 185
prudential rationality, 110–11
psychological causation
infants’ understanding of, 49, 50
without psychological mechanisms, 64–66
psychological vs. physical causality, 8
R-S theory, 101–2, 104, 105, 108–11
race as cause, 300
“radical conceptual change,” 331
rational agent theory, 328
rational analysis model, Anderson’s, 183
rational causation, 61, 64
reaching events, 272
real essences, 190
reasoning, 87, 94–97. See also causal reasoning
“precausal,” 8, 262
recency vs. primacy effects in causal induction, 281–84
Reichenbach, H., 351 nn.12. See also R-S theory
Reichenbach asymmetry, 247–50
Reichenbach dyads, 248–50
representativeness judgments, stability of, 183
Rescorla-Wagner (R-W) account of causal learning, 10–12, 119
Rescorla-Wagner (RW) model, 281–83
retrospective inferences made by younger children, 144
science, 67, 101, 264
scientific theories, 67, 70, 80
scoring, 123, 124
“screening off,” 5
screening-off reasoning, 141
search strategies, 257–58
second-order features, 176
second-order prototype (SOP) category similarity function, 176, 179–82, 185
self and other, differentiation between, 40
setting priors, problem of, 150
sex as cause, 300
Shultz, T. R., 256–57
similarity functions, three, 174–80
Skinnerian theory of infant imitation, 38
Sloman, Steve, 294–96
social effects, interventions, and invariance, 294–300
soft intervention, 64
spatial cognitive maps, 274–75
spatial relations, 71
specific theories, 305–7
spurious vs. causal relationships, 104
stability across background conditions, 106–7
statistical and causal structure, 249
statistical data
causal structure and, 157, 248
vs. temporal order, 159–61
statistical jokes, 295
statistical learning, 139, 144–46, 150
across modalities, 146–47
how it informs our understanding of causal learning, 147–49
new directions for integrating causal learning with, 149–50
problem of constraining, 149–50
(p.358) statistical regularity, 140
and causal knowledge, 140–41
“stickball machine,” 76–79
strength, 73, 105–6
focus on, 156
stressful life events and depression, 64–65
structural predicates, 336
structure, 337
vs. strength, 154–56
Suppes, P. See R-S theory
“supramodal” framework, 40
syntactic comprehension, 309
task complexity, 283–84
temporal order, 159, 162, 164, 165
vs. intervention, 163
vs. statistical data, 159–61
tertiary relationship, 32
theoretical coherence, 201–2
theories, 67, 70, 304, 306. See also specific topics
theory theory, 67, 68, 72, 77, 80, 262, 274
thought insertion, 62
tools, learning to use, 43–45
total cause (TC), 20–25
total effect, 108
total evidence, principle of, 249
two-cause trials, 141
undoing, 91–92
unexplained effects, 285
universal grammar (UG), 310, 311, 314
universal theory (UT), 311, 314
unobservables, 31
unobserved causes, inferring the existence of, 76–77
vision science, 7–8
“wrong variables” analysis, 32
zero sum competitions, 298–300