A Review of Ruth Byrne, The Rational Imagination:
How People Create Alternative to Reality

Carol Slater
Department of Psychology
Alma College
Alma, Michigan 48801-1616
USA
cslater@alma.edu
© Carol Slater, 2006

PSYCHE 12 (1), January 2006
REVIEW OF: Byrne, Ruth M. J. 2005. The Rational Imagination: How People Create
Alternatives to Reality. Cambridge MA: The MIT Press, 472 pp, ISBN: 0262025841.

Introducing The Rational Imagination, Ruth Byrne tells us that rational thought has
turned out to be more imaginative than cognitive scientistssupposed, andmore to
the point herethat [I]maginative thought is more rational than scientists imagined
(xi). It would be unwise to take this mini-manifesto (or the books title) too seriously. The
claim to which Byrne actually gives sustained attention is less philosophically sexy and
more solidly empirical. This book is primarily concerned with experimental evidence
(much of it Byrnes own) in support of the thesis that the particular counterfactual
conjectures people entertainIf Mary had asked Peter to pick the peppers, he would
have picked the peppersare governed by the same small set of psychological
principles that influence inferential reasoning about themPeter didnt pick the
peppers? Well, then, it stands to reason that Mary didnt ask him to (214-215). Byrne
conjectures that this same small set of principles might also help in understanding how
people creatively generate new members of a category (190191), interpret novel phrases
like cactus fish (192193), and solve insight problems (194-195). By contrast, Byrnes
discussion of criteria for the rationality of counterfactual thought comes close to the end
of the book and is notably modest and tentative. Perhaps counterfactual thought counts as
rational if it is capable of producing the best judgments; perhaps the best counterfactual
judgments are those that strike us as most plausible; perhaps plausibility is a hallmark of
rationality because it is grounded in recognition of fault lines in reality (208212). On
the other hand, perhaps not. Counterfactual thoughts that paralyze people with regret are
often compellingly plausible. (Try to deny If only I had looked in on the baby, I would
have noticed that something was wrong.) Despite their plausibility, Byrne characterizes PSYCHE: http://psyche.cs.monash.edu.au/
PSYCHE 2006: VOLUME 12 ISSUE 1
2
such dysfunctional counterfactuals as irrational. Perhaps this can be harmonized by
the competence/performance distinction; perhaps a canny reader would be better advised
to settle for the psychology
.
Byrnes approach in The Rational Imagination is unabashedly pedagogic. After
pointing out that logicians call an Ifthen premise a material conditional and
distinguishing these from biconditionals, Byrne asks the reader, What did you think
about when you understoodIf Alice went to the stables she rode Starlight? and
instructs them, Take a moment to list what is possible (and what is impossible) given the
truth of the conditional (19). Similarly, after an explanation of subjunctive mood, the
reader is asked, Do you think that if Oswald had not killed Kennedy someone else
would have? (32). Readers who anticipate that such prompts will be distracting might
prefer
to
consult
the
award
winning
Mental
Models
website
http://www.tcd.ie/Psychology/Ruth_Byrne/mental_models/ for references to other of
Byrnes publications

The Rational Imagination takes us through the following steps (29, 215).

[1] Human reasoning is rational;
[2] Human reasoning depends on the mental representation of possibilities;
[3] Which possibilities are selected for representation in reasoning is guided by a
set of principles;
[4] The set of principles that guide the possibilities people think about when they
reason also guide their imaginative thoughts (215);
[5] Imaginative thought is rational.

In Chapter 2, Byrne addresses [1] in the context of deductive reasoning. Rational
thought is thought that leads to reasonable conclusions (29); carried out correctly,
deductive reasoning yields a conclusion that is eminently reasonableone that is not
just plausible or possible [but that] must be true, if the factors it is based onare true
(15, stress in the original). People are deductively rational because they can appreciate a
simple semantic principle: an inference is valid if there are no counterexamples to it
(17). Systematic flaws in deductive reasoning pose no challenge to the existence of such
competence: like lapses from grammatically, they reflect only constraints on
performance. Readers will recognize [2] as the central axiom of mental model theory
(henceforth, MM) that provides the conceptual framework for the book. According to
MM, people understand a claim by imagining states of affairs consistent with its truth
true possibilitiesand test deductive inferences for validity by canvassing these
imagined possibilities for counterexamples.
For [3], Byrne identifies a set of principles that influence which possibilities will
be represented in reasoning, first, about factual conditionals and then about
counterfactual ones. People initially keep in mind just a few of what they take to be true
possibilitiesperhaps only one; some ideas, however, require that two possibilities be
represented; counterfactual thought involves the representation of two possibilities, one
identified as imagined or conjectured, the other as presupposed or factual; PSYCHE: http://psyche.cs.monash.edu.au/
C. Slater: Review of The Rational Imagination
3
counterfactuals are easier to generate when two alternatives are represented from the start
(40); in understanding obligations, people think about forbidden possibilities as well as
permitted ones; representations of possibilities include information about the temporal
order of events in the world (161).

Chapters 3 through 7 utilize the same groundplan to support [4]. First, some
puzzling facts about counterfactuals are displayed, e.g., that counterfactual thoughts are
more likely to undo an action than to alter an inaction, as assessed by judgments about
which would result in greater regret or satisfaction. Next, Byrne turns to studies of studies
of counterfactual reasoning for evidence of selection principles in operation therefor
example, studies that compare the inferences people draw from counterfactual as opposed
to factual premises, indicating that people represent two possibilities when they
understand a counterfactual conditional and only one in the case of a factual
conditional. Finally, Byrne discusses how, in conjunction with corollaries, these
principles can explain the candidate phenomenon. A relevant corollary here is that people
are more likely to represent actions than inactions with two possibilities.

Inactions are mentally represented more economically than actionsThere is no
change in state for inactions, and so the preaction and postaction possibilities
remain the sameThere are more things to keep in mind when someone does
something than when they do nothingBecause people keep in mind more
information for actions than for inactions, they can mentally change actions more
easily than inactionsPeople can readily imagine a counterfactual alternative to an
intentional action because they have envisaged two possibilities when they
understand the action(pp. 53-54).

Chapter 4, Thinking About What Should Have Happened, is likely to be of particular
interest to readers who have been following the debate about so-called content effects
on Wason-type selection tasks. Byrne firmly rejects explanations that invoke either
familiarity with permission schemas or the operation of mental modules specialized for
reasoning about social contracts. She offers, instead, an account in terms of MM and the
dual representation of forbidden and permitted possibilities (9296), concluding that

[I]f any sort of thinking did evolve[it] is the ability to think about certain
possibilities readily. Perhaps what evolved is the ability to understand obligations
by thinking about not only the permitted possibility but also the forbidden
possibility. This suggestionalso explains why people keep these two possibilities
in mind for indicative conditionals placed in contexts that elicit forbidden
possibilities (96).

Chapter 5 takes up complex relationships between counterfactuals and causal
thinking, asking, among other things, why an aspect of reality that has been targeted for
alteration in the antecedent of a counterfactual will not necessarily be accorded causal
clout with regard to the consequent. Chapter 6 continues the discussion of causal PSYCHE: http://psyche.cs.monash.edu.au/
PSYCHE 2006: VOLUME 12 ISSUE 1
4
reasoning with consideration of even if semifactuals that deny causal influence.
Chapter 7 addresses observations that people are more likely to alter the last event in a
series than an earlier one, as shown by our tendency to blame a player more when she
misses a free throw i