LiteratureToM Resub Copyedited

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/321257807
Effects of exposure to literary narrative ﬁction: From book smart to street

smart?
Article in Scientific Study of Literature · November 2017

DOI: 10.1075/ssol.7.1.06dem
CITATIONS READS
17 1,557
5 authors, including:
Hannah De Mulder Frank Hakemulder

Leiden University Utrecht University
21 PUBLICATIONS 71 CITATIONS 55 PUBLICATIONS 993 CITATIONS
SEE PROFILE SEE PROFILE
Rianne van den Berghe Fayette Klaassen

Windesheim University 16 PUBLICATIONS 148 CITATIONS
22 PUBLICATIONS 498 CITATIONS
SEE PROFILE
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
L2TOR - Second language learning with social robots View project
“Empirical Study of Literature Training Network” (ELIT) View project
All content following this page was uploaded by Hannah De Mulder on 29 May 2019.
The user has requested enhancement of the downloaded file.

EFFECTS OF EXPOSURE TO LITERARY NARRATIVE FICTION
Please cite as:
De Mulder, H. N. M., Hakemulder, F., van den Berghe, R., Klaassen, F., & van Berkum, J. J.
(2017). Effects of exposure to literary narrative fiction. Scientific Study of Literature, 7(1),
129-169. https://doi.org/10.1075/ssol.7.1.06dem
1
Running head: EFFECTS OF EXPOSURE TO LITERARY NARRATIVE FICTION
Effects of exposure to literary narrative fiction: From book smart to street smart?
Hannah De Mulder1*, Frank Hakemulder2, Rianne van den Berghe1, Fayette Klaassen3 & Jos
van Berkum1
1
UiL OTS, Utrecht University, The Netherlands
2
Department of Media and Culture, Utrecht University, The Netherlands
3
Department of Methodology and Statistics, Utrecht University, The Netherlands
*Corresponding author
Van Wijkplaats 4
2311 BX Leiden
The Netherlands
Contact: h.n.m.de.mulder@hum.leidenuniv.nl
Hannah De Mulder is now at LUCL, Leiden University, The Netherlands
Acknowledgements
We would like to thank the Netherlands Organisation for Scientific Research (NWO) for its
financial support (Vici grant #277-89-001 and Gravitation grant #024.001.003 to JvB). We
are also very grateful to Herbert Hoijtink for his help with the statistical analyses, to Ella
Bosch and Marjan Bezemer for their assistance in conducting the study and to Chris van Run
for his help with programming the tasks.
2
Abstract
Literary narrative fiction may be particularly effective in enhancing Theory of Mind (ToM),
as it requires readers to contemplate author and character intentions in filling the literary
‘gaps’ that have been suggested to characterise this fiction type. The current study
investigates direct and cumulative effects of reading literature on ToM using confirmatory
Bayesian analyses. Direct effects were assessed by comparing the ToM skills of participants
who read texts that were manipulated to differ in the amount of gap filling they required.
Cumulative effects were assessed by considering the relationship between lifetime literary
fiction exposure and ToM. Results showed no evidence for direct effects of reading literature
on ToM. However, lifetime literary fiction exposure was associated with higher ToM,
particularly cognitive ToM. Although reading a specific piece of literary fiction may thus not
have immediately measurable effects on ToM, lifetime exposure to this fiction type is
associated with more advanced ToM.
Keywords: literary narrative fiction, Theory of Mind, literary gaps, literariness, author
recognition test, Bayesian analyses
3
Effects of exposure to literary narrative fiction: From book smart to street smart?
Narrative Fiction and Understanding Others’ Mental States
“On the morning the last Lisbon daughter took her turn at suicide—it was Mary this
time, and sleeping pills, like Therese—the two paramedics arrived at the house knowing
exactly where the knife drawer was, and the gas oven, and the beam in the basement from
which it was possible to tie a rope.” (p. 1; Eugenides, 1993)
The quote above is the first sentence of Eugenides’ novel The virgin suicides and we
are immediately sucked in: Who are Mary and Therese? Who are these other daughters who
presumably died by stabbing, gassing and hanging themselves? And why on earth would
multiple daughters in one family all commit suicide? This example illustrates a hallmark of
narrative fiction: It is about autonomous intentional agents and their (inter)actions. If we are
to understand the story, we have to try to make sense of the characters’ social interactions,
that is, we must engage in advanced social-cognitive processing (Mar & Oatley, 2008;
Zunshine, 2006). Social-cognitive processing is thus necessary in order to be able to
appreciate narrative fiction (Barnes, 2012), but various researchers have suggested that this
link also works the other way round: Reading narrative fiction might actually enhance our
ability to engage in social-cognitive processing (Mar & Oatley, 2008; Oatley, 1999a, 1999b;
Oatley & Mar, 2005; Sugiyama, 2001; see Mumper & Gerrig, 2016 for a meta-analysis). The
idea is that narratives give us unique insights into the mental lives of the characters that
populate the story world. Furthermore, narratives provide us with the opportunity to live
through the social situations that the characters in the narrative go through as if we were there
experiencing them ourselves.
4
Reading fictional narratives is thought to enhance this experience further. The fiction
reader knows that there are no moral obligations to help anyone in the real world and that
there is no risk of experiencing any negative real life social consequences. This knowledge
likely enables the reader to relax her defences and freely indulge in thoughts and feelings that
may be triggered by the narrative (Hakemulder, 2000; Keen, 2007; Koopman, 2016a;
Koopman & Hakemulder, 2015). In this sense, then, our ability to understand others’ mental
states, our Theory of Mind (ToM) skills, may be deepened by reading narrative fiction.
In line with this idea, recent correlational studies have provided evidence that there is
a positive relationship between cumulative exposure to narrative fiction across the lifetime
and ToM in adults (Djikic, Oatley, & Moldoveanu, 2013; Fong, Mullin, & Mar, 2013; Mar,
Oatley, Hirsh, dela Paz, & Peterson, 2006; Mar, Oatley, & Peterson, 2009; Panero et al.,
2016; Stansfield & Bunce, 2014). Similar findings have also been reported for children (see
for instance Adrian, Clemente, Villanueva, & Rieffe, 2005; Aram & Aviram, 2009;
Gamannossi & Pinto, 2014; Ratner & Olver, 1998). Whilst these correlational studies cannot
be used to determine what the causal direction of the relationship is, various intervention
studies have suggested that exposure to narrative fiction improves mental state understanding.
For instance, reading a work of narrative fiction (as compared to reading a non-fiction text)
was found to increase the self-reported tendency to take others’ psychological viewpoint (at
least for those low in the personality trait openness, see Djikic et al., 2013). Furthermore,
various studies have demonstrated that performance on objective measures of mental state
understanding (i.e., measures that assess ToM ability directly instead of relying on self-
report) are also enhanced after exposure to narrative fiction as compared to exposure to non-
fiction (Black & Barnes, 2015; Kidd & Castano, 2013; Kidd, Ongis, & Castano, 2016; Pino
& Mazza, 2016).
5
What Are the ‘Active Ingredients’ of Narrative Fiction That Enhance ToM?
Previous research thus suggests that exposure to narrative fiction is beneficial for
ToM. What is not clear, however, is what exactly the ToM-enhancing ‘active ingredients’ in
narrative fiction might be. In their study, Fong et al. (2013) suggest that specific kinds of
narrative fiction may be more effective than others in enhancing ToM. They investigated the
relationship between exposure to various different narrative fiction genres (Domestic Fiction,
Romance, Science-Fiction/Fantasy and Suspense/Thriller) and ToM task performance.
Although the genres of Domestic Fiction, Romance and Suspense/Thriller were all related to
better ToM scores (and overall exposure to fiction was also positively related to ToM), only
exposure to Romance was still significantly correlated with ToM scores once relevant control
variables (e.g., age, gender and exposure to non-fiction) were taken into account. Potentially,
then, the fact that Romance fiction focuses so clearly on interpersonal relationships may
stimulate its readers to ponder on social interaction more generally and thereby come to a
greater understanding of other people’s mental states. Although this study suggests that
specific types of narrative fiction may be more clearly associated with ToM enhancement
than others, a limitation of this work is that it is correlational in nature. It is thus possible that
those who already have a better developed ToM prefer to read Romance fiction, instead of
exposure to Romance fiction being responsible for the ToM enhancement.
Kidd and Castano (2013) also investigate the possibility that specific kinds of
narrative fiction may be more effective than others in enhancing ToM. However, instead of
contrasting different kinds of genre fiction (as Fong et al., 2013, did), they considered
whether the literariness of a text could be linked to ToM enhancement. In their intervention
study, Kidd and Castano (2013) assessed the ToM abilities of groups of participants who had
either read literary narrative fiction, non-literary narrative fiction (popular fiction), expository
non-fiction or nothing immediately prior to a ToM assessment. The results suggest that
6
literary texts directly enhance ToM ability, whereas non-literary texts and expository non-
fiction texts do not. These results were replicated (and shown to hold for a new set of literary
and popular texts) in Kidd et al. (2016). Furthermore, a recent conceptual replication of this
study by Pino and Mazza (2016) found similar results and thus also suggests that the
literariness of the text is relevant in providing the reader with direct ToM gains. A
correlational study by Kidd and Castano (2016) also suggests that exposure to literary
narrative fiction over the lifetime is associated with higher ToM scores, whereas this was not
the case for exposure to popular narrative fiction. However, it should be noted that these
findings are debated in the literature. They are in contrast with the finding by Fong et al.
(2013) that Romance fiction is positively correlated with ToM (given that Romance is not
generally considered to be an instantiation of literary fiction). Furthermore, it is also not
entirely clear to what extent these findings can be replicated, as there are various intervention
studies that do not find a special relationship between exposure to literary narrative fiction
and ToM. For instance, both Panero et al. (2016) and Dijkstra et al. (2015) did not find direct
ToM enhancement for those participants who had read literary narrative fiction as compared
to readers of other text types. Dijkstra et al. (2015) focused on the comparison between
literary vs. non-literary narrative fiction and did not find any ToM differences. Panero et al.
(2016) compared literary texts with both non-literary narrative fiction and non-fiction and
similarly did not find any ToM differences between readers in the various conditions
(although Kidd & Castano, 2017, debate the validity of these conclusions in a commentary on
the paper by Panero et al., 2016, and provide a reanalysis of the data that suggests that the
Panero et al., 2016, findings are in fact in line with the results of Kidd and Castano, 2013).
There is thus debate in the literature on whether exposure to literary narrative fiction
leads to direct ToM gains in the reader. But if we assume for the moment that literariness is
indeed related to readers’ ability to understand others’ mental states, this raises the question
7
why specifically literary narrative fiction might be more effective than other types of
narrative fiction in enhancing our understanding of others. Potentially, literary narrative
fiction provides the reader with a particularly rigorous social-interaction training, as this type
of text requires deep psychological processing of authors’ intentions and the characters’
subjective experiences in order for full understanding to be achieved. Furthermore, literary
texts tend to be more ambiguous than expository or non-literary narrative texts (see Koopman
& Hakemulder, 2015). In contrast, non-literary narrative fiction is considered to provide a
more internally consistent and predictable portrayal of the context and characters (see Gerrig
& Rapp, 2004; Van Peer, 1986) and thus might not provide the reader with a ‘ToM workout’
that is rigorous enough to have immediately measurable effects (or, indeed, any effects at all).
It is likely, then, that reading literary narrative fiction requires more profound psychological
‘work’ on the part of the reader as compared to non-literary narrative fiction and thus
enhances ToM, both directly after exposure and as a cumulative effect of exposure over the
lifetime (see also Kidd & Castano, 2016, for discussion of this issue).
That this interpretation of previous findings may be correct is supported by a study
conducted by Peskin and Astington (2004) on the effects of narrative fiction exposure on the
ToM abilities of young children. Children who were read a book that portrayed many
situations in which an understanding of others’ mental states was relevant, but that did not
make these mental states explicit in the text itself, subsequently did better on measures of
ToM than those who were read a book with the same storyline, but with rich mental state
language. Likely then, children who were not exposed to explicit mental state language in the
text had to work harder in order to understand the storyline and hence came to a deeper
understanding of others’ mental states than those who had received more support from the
text itself. Having to engage in deep cognitive processing to be able to understand a given
8
story character’s mental states thus seems to be more beneficial in developing ToM than
being provided with the content of others’ mental states explicitly.
This process may not only apply to children, but to adults as well, as is suggested by
the work of Kidd and Castano (2016). In this study it was found that literary texts contain
more markers of reflective function than genre fiction texts. This characteristic of literary
texts was found to mediate the effect of literary fiction on ToM performance. In line with the
findings reported by Peskin and Astington (2004), these markers of reflective function do not
reduce to the presence of mental state terms in a text, but instead relate to how mental states
are discussed (e.g., although I was angry, now I am sad receives a relatively low reflective
function rating, as the speaker only reports his emotions at different times, an utterance like I
felt really angry and then it changed to sadness receives a much higher rating as the speaker
demonstrates an awareness of how their emotional response to something transformed over
time). Indeed, further analyses reported in Kidd and Castano (2016) demonstrated that there
was no difference between literary and genre fiction texts regarding the presence of words
related to social and cognitive processes and these variables were not found to mediate the
relationship between text type and ToM scores. There is thus some suggestion in the
literature that the level of psychological ‘work’ that a reader has to engage in, which is
typically higher in literary narrative fiction than popular fiction, may be important in
determining whether or not a text will enhance ToM both immediately after reading and after
extended exposure.
The current study aims to investigate whether it is indeed the case that exposure to
literary narrative fiction, given its potential as a particularly rigorous ToM workout, enhances
ToM in adults. This issue will be considered at the direct and the cumulative level,
investigating both how reading specific pieces of literary narrative fiction affects ToM (direct
9
effect) and how exposure to this type of narrative fiction over the lifetime relates to ToM
skills (cumulative effect).
Literary Narrative Fiction and ToM: The Relevance of ‘Gaps’
In investigating whether there are direct effects of literary narrative fiction on ToM
skills, the current study focusses on the presence of literary ‘gaps’ in a text (Iser, 1978, 1988).
In Iser’s conceptualization of information gaps it is assumed that these are characteristic of
literature. In literary texts, we find gaps that have to be (more or less) filled in order to allow
for a full (or fuller) comprehension of the narrative by the ‘ideal’ or implied reader. The text
invites, or appeals to, the reader to be creatively involved in filling in these gaps (cf. Iser,
1975). Time lapses, discontinued sub-plots and open endings are examples of narrative
devices that may give rise to literary gaps as they stimulate readers to fantasize about what
happened in the meantime and to postulate character behaviour on the basis of what they
already know about the character. Similarly, texts that provide conflicting perspectives of two
or more narrating characters also provide many gaps to be filled (regarding, for instance, the
narrators’ morality and reliability in their telling of events). Another example of gap filling
relates to the extent to which the social situation that readers are introduced to is clarified in
the text (cf. Van Peer, 1989). Readers in popular fiction are typically informed directly from
the start who is who (e.g., by referring to characters by naming them when introduced),
which thoughts and utterances are to be attributed to which character and explicitly defining
what the relationship between the characters is. However, content analysis of a corpus of
popular and literary narratives (Van Peer, 1989) suggests that in literary texts such aspects are
often left ‘open’ (as the opening sentence of the literary novel by Eugenides at the beginning
of this article illustrates). It is this latter type of gap filling that will be manipulated in the
current study, as gaps of this nature most clearly require readers to contemplate characters’
10
thoughts and social relationships in order to be able to make sense of the narrative. Following
Kidd and Castano (2013), who suggest that the presence of literary gaps in the narrative
structure may be important in explaining the relationship between literature and ToM, our
assumption is that exposure to a text that is rich in these kinds of gaps will have a more
pronounced direct effect on ToM than texts with fewer (or no) gaps. In order to investigate
whether this is indeed the case, in the current study readers were provided with a literary
narrative fiction text (one of the texts used in Kidd & Castano, 2013) or a non-fiction text.
The literary text was either read in its original form (for maximal levels of required gap-
filling) or in one of two adapted forms in which the amount of psychological work that was
required was either somewhat or considerably reduced. For instance, readers of the adapted
texts were immediately informed of the identity of one of the central figures in the story,
whereas the original text does not make this information clear until much later in the text (see
the Method section for more information on the text manipulations and further examples). A
final group of readers received an expository non-fiction text, such that effects of reading a
text for which literary gap filling was not required could also be assessed. After exposure to
one of these four texts, participants were given various ToM measures, such that ToM
differences between the groups could be determined.
Aside from assessing whether reading a specific piece of literary narrative fiction
would have a direct effect on ToM, the current study also investigated whether exposure to
literary narrative fiction over the lifetime would have a cumulative effect on ToM. Given that
previous correlational studies have found exposure to narrative fiction in general over the
readers’ lifetime to be associated with ToM (e.g., Panero et al., 2016) and that literary
narrative fiction in particular may be linked to higher ToM (e.g., Kidd & Castano, 2013,
2016; Kidd et al., 2016), our assumption was that more exposure to literary narrative fiction
over the lifetime would result in more advanced ToM skills. In order to investigate this
11
question participants were given a test that assessed their exposure to both literary and
popular narrative fiction (based on author recognition, see Method section). In this way, the
current study can assess the effects of literariness on ToM ability both in a direct sense
(depending on what particular text a participant has been exposed to) and cumulatively
(depending on the extent of lifetime literary narrative fiction exposure).
This design thus enables us to contribute to current knowledge on the relationship
between narrative fiction and ToM in various ways. In the first place, the study will provide
additional experimental evidence for the presence or absence of direct effects of reading
literary narrative fiction on ToM. Although some previous studies have found evidence for
this relationship (Kidd & Castano, 2013; Kidd et al., 2016; Pino & Mazza, 2016), other
studies have not (Dijkstra et al., 2015; Panero et al., 2016). Our study aims to add to this
debate not only by comparing the ToM skills of readers of literary narrative fiction vs. other
types of texts (as previous studies have done), but also by focusing on one particular
characteristic of literary narrative fiction, the presence of literary gaps, that is likely to be an
‘active ingredient’ in enabling direct ToM gains. Secondly, the current study investigates
whether cumulative ToM gains are present for those who have higher lifetime exposure to
literary narrative fiction. Whereas previous studies have often focused on cumulative ToM
effects related to general exposure to narrative fiction (e.g., Djikic et al., 2013; Mar et al.,
2006, 2009; Panero et al., 2016), the current study is one of the few studies (together with
Kidd & Castano, 2016) that assess whether this relationship might also hold for, or even be
specific to, literary narrative fiction.
Competing Hypotheses Regarding Effects of Narrative Fiction Exposure on ToM
The current study employs a confirmatory Bayesian approach of data analysis, which
entails that prior to data analysis various competing hypotheses regarding the possible
12
outcomes of the study are formulated in a very precise manner. Which specific competing
hypotheses are chosen is based on the outcomes of previous studies and on current theorising
on the matter (Hoijtink, 2012). The formulated hypotheses are then evaluated against the data
in order to determine which hypothesis receives the most support in the current dataset. The
Data Analysis section provides more information on this analysis approach and a more
precise formalisation of the hypotheses, but an overview of the hypotheses and the basic
assumptions behind them are detailed here.
To investigate direct effects of exposure to (literary) narrative fiction on ToM ability
four competing hypotheses were formulated. Our first hypothesis is based on Kidd and
Castano’s (2013) results and states that the more literary the text that the reader has been
exposed to, the greater its effect will be on ToM ability immediately after exposure. Reading
the original literary text will thus be associated with the highest ToM scores, followed by the
two adaptations (in order of literariness) and the expository non-fiction text. Our second
evaluated hypothesis is based on a more extreme interpretation of Kidd and Castano’s
findings. Given that only literary texts were found to enhance ToM ability in their study, it
could also be the case that any manipulation that decreases a text’s literariness immediately
‘devalues’ it to such an extent that it no longer qualifies as a literary text and thus no longer
has any immediately measurable effects on ToM ability. If this is indeed the case, then the
highest ToM scores will be obtained by the readers who were exposed to the original version
of the literary text, but there will be no difference in ToM scores for the readers who were
exposed to any of the other texts. Our third hypothesis is based on the body of literature that
has not made a specific distinction between different types of narrative fiction, but that
suggests that narrative fiction in general enhances ToM more than non-fiction (e.g., Mar et
al., 2006, 2009). Following this hypothesis, we would expect to see higher ToM scores for
the readers of the narrative fiction texts as compared to the expository non-fiction text, but no
13
differences between scores depending on the literariness of the narrative fiction text that had
been read. The final hypothesis that was formulated is based on the results of Panero et al.
(2016) and states that reading literary narrative fiction is not associated with direct ToM
gains, whatever the level of literariness of the text. If this is the case, we would not expect to
see any differences in ToM between the various reader groups in the study.
To investigate cumulative effects of exposure to (literary) narrative fiction on ToM
ability, the current study also formulated hypotheses regarding effects of lifetime exposure to
narrative fiction, both in general and as regards literary and popular narrative fiction in
particular. For general exposure to narrative fiction (i.e., exposure to literary and popular
narrative fiction combined), three competing hypotheses were formulated: Narrative fiction
exposure could be positively, negatively or not related to ToM ability. We expected there to
be a positive relationship between the two given results of previous studies (Djikic et al.,
2013; Mar et al., 2006, 2009; Panero et al., 2016). The two logical opposites of this
hypothesis were investigated in order to determine how well this hypothesis would fare
against them.
The final three competing hypotheses were formulated with regards to exposure to
literary narrative fiction over the lifetime. If literary narrative fiction is of primary
importance in enhancing ToM ability (as suggested by Kidd & Castano, 2013, 2016), we
would expect to see scores on lifetime literary narrative fiction exposure to be positively
correlated with ToM and more strongly so than lifetime exposure to popular narrative fiction.
To determine the strength of this hypothesis, we compared this hypothesis to two clear
counterhypotheses. The first of these two counterhypotheses evaluated the possibility that
lifetime literary and popular fiction exposure would not differ in their relation to ToM ability
(in line with what would be expected if the outcomes of the intervention studies by Dijkstra et
al., 2015, and Panero et al., 2016, also hold for lifetime exposure). The second, more extreme,
14
counterhypothesis evaluated the possibility that there would be no relation between lifetime
literary exposure and ToM at all (in direct contrast to the claims by Kidd & Castano, 2013,
2016).
Method
Participants
121 participants (90 women), recruited from a large pool of participants at the
University of Utrecht in the Netherlands, took part in this study in exchange for a small fee (5
euros). Participants were non-dyslexic native speakers of Dutch between 18 and 30 years old
(M = 21;8, SD = 2.9). By far the most were students, but some participants indicated having
other occupations.
Materials
Four different texts were used (full translated versions of the texts can be found in the
Appendix): An original literary text, two adaptations of this same text (differing in the
amount of gap-filling they required) and an expository non-fiction text. The literary text that
was used was the short story A chameleon by Chekhov (1884/1978). This text was chosen as
it was also part of the study by Kidd and Castano (2013) and it had a Dutch translation
available (Chekhov, 1884/2014). The material was manipulated such that it would be possible
to assess to what extent the presence of literary gaps would be associated with ToM gains. A
characteristic of the story A chameleon is that it remains implicit which character is
responsible for the many instances of direct discourse that are present in the text. Figuring out
who said what is thus a crucial element of the original version of the story. Furthermore, the
story contains various instances of oblique reference (i.e., it is not immediately clear to
whom/what particular pronouns refer) and the social relations between characters are also not
15
always clarified in the text (see Table 1 for examples; changes are rendered in boldface). Our
assumption was thus that the original text would place a greater burden on readers to fill these
literary gaps and thereby provide them with a better ToM workout than manipulated versions
of this text in which some or most of these gaps had been filled.
The three literary text conditions consisted of the story A chameleon in its full original
form (consisting of 1216 words), in a version in which the text had been manipulated such
that only the prominent gaps in the story were filled (1233 words, limited manipulation
version) or in a version in which many more of the gaps were filled (1214 words, extensive
manipulation version). In the limited manipulation version instances of direct discourse were
provided with sources and initially oblique references were clarified immediately. The
extensive manipulation version contained all the changes present in the limited manipulation
version as well as further changes that were intended to make it even easier for readers to
understand the social situation of the story (and hence character motives and emotions) by
removing all possible obstacles to this understanding. In the first place, the story was adapted
such that it would fit in the modern Dutch context that was familiar to the participants instead
of the older Russian context of the original. This entailed that all Russian names were
changed to Dutch names, and that Dutch professions, authority figures and prototypical
hierarchical orders were used instead of Russian ones. We assumed that these ‘translations’ to
the more familiar context of the readers would entail that they would have to put less effort
into understanding who was who and how everyone was related to each other. Secondly,
parts of the text were deleted from the story if we thought that they may give rise to literary
gap-filling without being directly relevant for plot comprehension. To give an example: At
the beginning of the story a parcel is prominently introduced, but it plays no role in the rest of
the story. As readers might be inclined to spend psychological work on guessing what the
author’s intention might have been in introducing this parcel, this information was deleted
16
from the text. By manipulating the text in these ways, we assumed that readers of the
extensive manipulation version would be provided with fewer opportunities to engage in
psychological ‘work’ than readers of the original text and the limited manipulation version.
Table 1
Examples of the original text and the two manipulated versions
Text type Text example
Original “Shut your mouth!”
“No, that’s not the General’s dog,” says the policeman, with profound
conviction, “the General hasn’t got ones like that. His are mostly
setters.”
“Do you know that for a fact?”
Limited manipulation “Shut your mouth!” Otsjoemelov orders.
“No, that’s not the General’s dog,” says policeman Jeldyrin with
profound conviction, “the General hasn’t got ones like that. His are
mostly setters.”
“Do you know that for a fact?” Otsjoemelov asks.
Extensive manipulation “Shut your mouth!” Jansen orders.
“No, that’s not the chief commissioner’s dog” says policeman De
Vries with profound conviction, “The chief commissioner hasn’t got
ones like that. His are mostly setters.”
“Do you know that for a fact?” Jansen asks.
Note. Differences as compared to original text are rendered in bold face.
The expository non-fiction text that was used was a slightly adapted version of an
article from the Dutch edition of the National Geographic magazine (Van Beem, 2015). This
article provided information on the biology and habitat of chameleons. Although the original
17
length of the expository non-fiction text was close to that of the literary fiction text, the
adapted version was somewhat shorter (985 words), as the more narrative and emotional
aspects of the text were filtered out such that the contrast between the two text types was as
great as possible.
Measures
Lifetime fiction exposure. Lifetime exposure to fiction was assessed using an Author
Recognition Test (ART, Stanovich & West, 1989) in which participants have to indicate
which author names they recognise from a large list of names. Although the ART does not
assess lifetime fiction exposure in a direct sense (by asking respondents directly about the
frequency and duration of their fiction reading habits), this measure is commonly interpreted
as a general measure for print exposure. It has been found to be a good predictor for other
measures related to print exposure such as reading skills (Mol & Bus, 2010) and exposure to
fiction (West, Stanovich, & Mitchell, 1993). Furthermore, this test has the added benefit that
it does not run the risk of obtaining socially desirable responses in answering direct questions
regarding reading frequency/duration. Attempts to assess general print exposure have
sometimes included more than just the names of fiction authors (e.g., magazine titles, see
West et al., 1993), but the current study employs a version of the ART created by Koopman
(2015a, 2015b, 2016b) that focuses specifically on exposure to narrative fiction authors.
Given this more targeted focus, we suggest that this version of the ART will be better capable
of providing us with information on lifetime exposure to narrative fiction than using a more
general ART. Furthermore, Koopman (2016a) found that avid readers of literature scored
significantly higher on her ART than a control group did. Moreover, it is important for the
present purpose that this ART has been specifically designed with Dutch participants in mind,
and that it allows for a distinction between recognizing literary and popular authors. (Authors
18
classified as literary had received critical acclaim and nominations for literary prizes; authors
in the popular category had written best-selling novels that are relatively easy to read.)
Although we acknowledge that we have no direct validation of the relationship
between recognition of literary vs. popular author names and actual exposure to literary vs.
popular narrative fiction, we note that the ART as a measurement tool has been found to
relate to actual narrative fiction exposure and that previous studies that apply factor analysis
to the ART find a two-factor solution that can be broken down into a literary and a popular
fiction factor (see Kidd & Castano, 2016; Moore & Gordon, 2015). We thus consider it to be
a reasonable assumption that the version of the ART that we use will be informative
regarding exposure to literary and popular narrative fiction in our Dutch sample.
Participants were given a list consisting of 42 names. Thirty of these names belonged
to actual authors, 12 of the names were foils. Of the 30 author names, 15 belonged to literary
fiction writers and 15 to popular fiction writers; in both categories national (i.e., Dutch) and
international author names were present. Participants could thus score a maximum of 30
points in total for the ART (reflecting general exposure to narrative fiction); the literary and
popular subcomponents of the ART (reflecting exposure to literary and popular narrative
fiction) were each worth 15 points. Participants were told that they did not have to have read
the work of the author themselves in order to check the name, they just had to be sure that the
name belonged to an actual author. Guessing was discouraged by informing participants that
there were also made-up names in the list.
ToM. Dutch versions of two different measures of ToM were employed: The Reading
the Mind in the Eyes Task (RMET, Baron‐Cohen, Wheelwright, Hill, Raste, & Plumb, 2001)
and the Yoni task (Shamay-Tsoory & Aharon-Peretz, 2007). We used these tasks as previous
work by Kidd and Castano (2013) has shown that performance on both of these tasks is
19
enhanced after reading literary narrative fiction. In the RMET task, participants are shown the
eye region of a series of 36 different faces and are required to choose from four possibilities
(e.g., irritated, sarcastic, worried or friendly) which option best conveys the mental state that
is being signalled. Although some of the response options refer to an understanding of
cognitive mental states (e.g., contemplative, thoughtful), by far the most of the response
options refer to affective mental states (e.g., hostile, anxious, affectionate). The RMET thus
primarily measures participants’ ability to attribute affective mental states to others (i.e., it
assesses knowledge about others’ feelings) and thus is primarily a test of affective ToM.
The Yoni task assesses ToM in a broader sense than the RMET, as understanding of
both cognitive and affective mental states is assessed in equal measure. Furthermore,
understanding of these two types of mental states is assessed at both the first and second
order level. In each trial, participants see a cartoon face called Yoni in the centre of the
screen, a picture in each corner of the screen and a sentence at the top of the screen. This
sentence expresses Yoni’s cognitive or affective mental state in relation to one of the four
other pictures on the screen. The participant’s goal is to determine to which of the four
pictures the sentence refers, given the verbal and visual clues to Yoni’s mental state that are
present in the stimulus. In the assessment of cognitive mental states (25 trials in total with 13
first order trials and 12 second order trialsi), the facial expressions of all characters in the
stimulus and the verbal cue are emotionally neutral. In the first order cognitive trials, the four
pictures in the corners of the screen are all of objects and the participant has to use Yoni’s eye
gaze direction in order to determine which of these objects Yoni is thinking about (example
stimulus sentence: Yoni is thinking of…). In the second order cognitive trials, the four pictures
each represent another character together with an object. For these trials, the participant has
to demonstrate understanding of Yoni’s beliefs about these other characters’ beliefs and
desires (e.g.,: Yoni is thinking of the toy that …wants). Cognitive trials thus assess
20
participants’ understanding of Yoni’s thoughts (in the first order trials) and Yoni’s thoughts
as related to others’ thoughts/desires (in the second order trials) in an emotionally neutral
context. In the affective trials (24 in total, with 12 first order and 12 second order trials), the
facial expressions of the characters and the verbal cue are affective in nature (either positive
or negative). In the first order trials, participants have to use this affective information to
determine Yoni’s feelings regarding a particular object (e.g.,: Yoni loves …). In the second
order affective trials, again, the four pictures consist of other characters together with objects
and the participant has to demonstrate an understanding of Yoni’s emotions in regard to the
other characters’ emotions in order to score points on the task (e.g.,: Yoni loves the toy that …
loves). Affective trials thus assess participants’ understanding of Yoni’s feelings (in the first
order trials) and Yoni’s feelings as related to others’ feelings (in the second order trials) in an
emotionally loaded context. These 49 trials, requiring first and second order understanding of
cognitive and affective mental states, are the core component of the Yoni task and were the
focus of the analysis in the current study. In addition to these 49 core items, participants also
received 49 additional trials that were not analysed for this studyii. These trials consisted of
items assessing understanding of Yoni’s physical states (e.g., Yoni is close to…) and items
that assessed understanding of emotions relating to the fortunes of others such as envy and
gloating (e.g., Yoni envies…; see Shamay-Tsoory, Tibi-Elhanany, & Aharon-Peretz, 2007, for
more information on these conditions).
Procedure
Participants were tested individually in a testing booth with each session lasting
between 30-45 minutes. Upon arrival, each participant read and signed an informed consent
form that gave them general information about the study. On obtaining consent, participants
were given general instructions about the experimental procedure, and then started the
21
experimental session by reading the text that they had been assigned to (the text was provided
to them on paper). Participants were encouraged to take as much time as they wanted to read
the text and to reread it if necessary. After having read the text, participants were given the
RMET, the Yoni task and the ART to complete. Finally, the participants received a
questionnaire that asked them to report on their experience during the experiment as a whole
and whilst reading the text they had been assigned to. This questionnaire also served to
ensure that participants had paid sufficient attention to the text, as they were required to
answer three multiple choice text comprehension questions. If at least two out of these three
questions were not answered correctly, the participant’s data was removed from the analysis.
Participants were also asked whether they had read the text (or a text that was similar to the
text they had read) before, such that participants who were already familiar with the texts in
the study could be removed from analysis. After each task, the experiment leader entered the
booth to start the next task and to give instructions tailored to the upcoming task. After
completing the final questionnaire, the participant was paid and debriefed.
Data Analysis
The data analysis approach employed by this study relied on informative hypotheses
and Bayesian model selection (Hoijtink, 2012) using the software BIEMS (Mulder, Hoijtink,
& de Leeuw, 2012). A crucial aspect of this approach is that, prior to observation of the data,
various competing hypotheses are specified based on information obtained from previous
research and current theorising (Kluytmans, van de Schoot, Mulder, & Hoijtink, 2012; van de
Schoot et al., 2011; van de Schoot et al., 2014). These specific hypotheses are then compared
against each other (and not against the null hypothesis as is the case in classical frequentist
hypothesis testing, see Harlow, Mulaik, & Steiger, 1997). The Posterior Model Probability
(PMP) that is associated with each hypothesis is then used as an indication of the amount of
22
evidence that there is in the data for each particular hypothesis over other evaluated
hypotheses (see Results section more information on this concept).
Due to the fact that only pre-specified hypotheses are assessed, use of smaller sample
sizes is not generally problematic in this approach (whereas this can be a cause for concern in
classical hypothesis testing). That is, under a Bayesian approach, what is being evaluated is a
very specific set of questions: Is the data structured as specified under Hypothesis X, Y or Z?
Depending on the PMP values that are associated with each hypothesis, one hypothesis may
clearly be more supported than the others, or there may be no clear ‘winner’, in which case
the conclusion can be drawn that none of the specified hypotheses is any better than the
others in accounting for the data. The only way in which sample size is of influence is that
larger sample sizes may lead to stronger support for one of the hypotheses (but support can
still be computed for smaller sample sizes). This is in contrast with classical hypothesis
testing in which the underlying question is much broader: Are the values associated with the
variables of interest the same or do they differ in any of the multitude of possible ways in
which they could potentially differ from each other? In this case, a small sample size may
incorrectly lead to a non-rejection of the null hypothesis because the power is too low.
This confirmative approach thus allows for more precise testing of hypotheses even
with relatively small sample sizes, does not rely on arbitrary cut-off points and reduces the
chance of obtaining false positives and false negatives, and, as such, is the approach of choice
for this study.
Direct effects. Four competing hypotheses relating to the effect of text literariness on
ToM were devised (see also the Competing Hypotheses Regarding Effects of Narrative
Fiction Exposure on ToM section). Hypothesis 1-DE (i.e., the first Hypothesis regarding
Direct Effects) assumes that exposure to the original literary text will be associated with the
23
greatest gains in ToM and that decreasing the amount of psychological work necessary will
have a detrimental effect on these gainsiii:
Hypothesis 1-DE: Original > limited manipulation > extensive manipulation > non-fiction
Hypothesis 2-DE presents the more extreme version of Hypothesis 1-DE and states that only
exposure to the original literary text will enhance ToM, whereas any decrease in literariness
will no longer lead to ToM improvement:
Hypothesis 2-DE: Original > limited manipulation = extensive manipulation = non-fiction
Following Hypothesis 3-DE, exposure to any kind of narrative fiction text, be it manipulated
or not, will have a favourable effect on ToM scores:
Hypothesis 3-DE: Original = limited manipulation = extensive manipulation > non-fiction
The final hypothesis, Hypothesis 4-DE, formalises the notion that exposure to (literary)
narrative fiction texts is not related to ToM performance:
Hypothesis 4-DE: Original = limited manipulation = extensive manipulation = non-fiction
Cumulative effects. For the cumulative effects of exposure to general narrative
fiction over the lifetime three competing hypotheses were formalised (see Figure 1 for a
visualisation of the correlation procedure that is employed in these analyses). Hypothesis 1-
CE (i.e., the first Hypothesis regarding Cumulative Effects) postulates that there is a positive
24
correlation between exposure to narrative fiction and ToM. The other two hypotheses have
been formulated such that they counter this notion: Hypothesis 2-CE states that there is a
negative relationship and Hypothesis 3-CE claims that there is no relationship between
narrative fiction exposure and ToM.
Figure 1. Hypotheses cumulative effects for general narrative fiction exposure on ToM.
Note. + denotes a positive correlation, - a negative correlation and =0 denotes a lack of correlation between the
dependent and independent variables.
In addition to these three hypotheses relating to effects of cumulative exposure to
narrative fiction in a general sense, the final three hypotheses regarding cumulative effects
considered effects of exposure to popular and literary narrative fiction over the lifetime (see
Figure 2, for a visualisation of the regression procedure that is employed in these analyses).
Hypothesis 4-CE predicts that exposure to literary fiction will be positively related to ToM
(formalised as ß1 > 0 with ß1 denoting the standardised regression coefficient for literary
narrative fiction exposure in relation to ToM) and more strongly related to ToM than
exposure to popular narrative fiction (formalised as ß1 > ß2, with ß2 denoting the
standardised regression coefficient for popular narrative fiction in relation to ToM). This
hypothesis is contrasted with two hypotheses that go against this notion. Hypothesis 5-CE
specified the claim that literary narrative fiction exposure does not deserve a special status,
but instead that there is no difference between exposure to literary and popular narrative
fiction in their relation to ToM (formalised as ß1 = ß2). Finally, Hypothesis 6-CE states that
exposure to literary narrative fiction is unrelated to ToM (ß1 = 0).
25
Figure 2. Hypotheses cumulative effects literary vs. popular narrative fiction on ToM.
Note. ß1 and ß2 are standardised regression coefficients; ß1 denotes the relationship between exposure to literary
narrative fiction and Theory of Mind; ß2 denotes the relationship between exposure to popular narrative fiction
and Theory of Mind.
Results
Descriptive Statistics and Correlations
Prior to data analysis, data from two participants was discarded. One of these two
participants had taken an excessively long time to read the text (taking more than twice the
time that the other participants needed on average) and complained of severe concentration
problems during the experiment in the final questionnaire. The other discarded participant
failed the exclusion criterion of answering at least two out of the three basic text
comprehension questions correctly (as posed in the final questionnaire)iv. All the analyses
reported below are thus based on the remaining set of 119 participants (all between 18 and 30
years old; M = 21;8, SD = 2.9; 89 women).
Table 2 provides the descriptive statistics of the scores on the ToM assessments and
the ART, both overall and broken down per text condition. Note that the cells above the
dashed border present the descriptive statistics that are relevant for the assessment of direct
effects on ToM, while the cells below this border relate to the cumulative effects. The ART
scores are broken down by text condition such that potential group differences in lifetime
26
exposure to narrative fiction can be considered. Although the popular ART scores are very
similar between groups, there are larger differences between groups for the ART literary
scores (and thus also the ART overall scores). This potential confound will be dealt with in
the analyses presented in the following sections.
It should be noted that only a very small number of foils was chosen in the ART (M =
0.24; SD = .58). By far the most of the participants (81.5%) checked zero foils (95.8% of
participants checked zero or one foil) and no more than three foils were checked by any
participant. Given that the number of foils checked was not part of our pre-specified
hypotheses and this inspection of the data did not provide a clear motivation for us to do so in
additional post-hoc analyses, this variable was not considered in further analyses.
Table 2
Descriptive statistics ToM (RMET and Yoni) and lifetime exposure to fiction (ART)
Note: M = Mean; SD = Standard Deviation; RMET = Reading the Mind in the Eyes Task; ART =
Author Recognition Task. Maximum values: RMET = 36; Yoni = 49; ART overall = 30; ART
27
literary/ART popular = 15. Cells above the dashed border relate to the assessment of direct effects on
ToM; cells below this border pertain to cumulative effects.
Table 3 provides the correlations between the ToM tasks and the lifetime exposure
measures.
Table 3
Correlations between ToM (RMET and Yoni) and lifetime exposure to fiction (ART)
RMET Yoni ART Overall ART Literary
Yoni .26
ART Overall .03 .12
ART Literary .03 .18 .92
ART Popular -.01 .05 .92 .71
Note. RMET = Reading the Mind in the Eyes Task; ART = Author Recognition Task.
Effects of Text Literariness: Direct Effects
In the interpretation of the results in the following sections the Posterior Model
Probability (PMP) will be used to determine which of the tested hypotheses is most
supportedv. The PMP gives the probability that a particular hypothesis is the most supported
by the data given all the specific hypotheses that are evaluated, taking into account the
complexity and the fit of the hypothesis. If a particular experimental hypothesis thus has a
PMP of, for example, 0.64, this entails that the probability is 64% that preferring this
hypothesis over other evaluated hypotheses is the right choicevi. Importantly, there should be
no rules regarding what constitutes a sufficient value. As Rosnow and Rosenthal (1989)
28
already stated regarding the .05 rule to decide whether a p-value is significant: “Surely, God
loves the .06 nearly as much as the .05” (p. 1277). Any rule for the interpretation of the size
of PMPs is subject to the same criticism. The results are thus summarised and reported in this
section and their interpretation is reflected upon in the discussion section.
Table 4 provides PMPs for Hypothesis 1-4-DE in relation to performance on the
RMET and Yoni task. The shaded cells represent the hypothesis with the highest PMP (i.e.,
the model that is most supported by the data). Aside from the PMP values of the specified
hypotheses (i.e., Hypotheses 1-4-DE which consist of pre-specified constraints), Table 4 also
provides PMP values for the unconstrained hypothesis. This hypothesis does not have any
formulated constraints (it formalises the notion something is going on, but I don’t know what)
and is generally considered in addition to the pre-specified hypotheses in order to get an
indication of the quality of the pre-specified hypotheses (Hoijtink, 2012). If the unconstrained
PMP value is high, this suggests that the constraints that have been specified are ill-chosen
(because evidently they are not supported better by the data than a hypothesis that does not
impose any constraints). As can be seen from this table and the other tables presenting the
results, this is not a concern in the current study, as the most supported hypotheses have
clearly higher PMPs than the unconstrained hypotheses.
29
Table 4
Direct effects of text condition on ToM (RMET and Yoni)
Hypothesis RMET PMP Yoni PMP
1-DE: More literary better ToM 0.01 0.05
2-DE: Original text better ToM 0.03 0.06
3-DE: Any literature better ToM 0.01 0.03
4-DE: No differences 0.83 0.81
Unconstrained 0.11 0.04
The results presented in Table 4 demonstrate that there is very little support in the
data to suggest that reading narrative fiction of any type is associated with better performance
on the ToM tasks. Indeed, the PMPs of Hypotheses 1-3-DE are very low, whereas Hypothesis
4-DE is very high both for the RMET and the Yoni task (the chance that Hypothesis 4-DE is
the right choice is 83% in relation to performance on the RMET task and 81% in relation to
performance on the Yoni task). The descriptive statistics presented in Table 2 also underscore
this conclusion: The differences in RMET and Yoni average scores between participants that
had read different text types are minimal. Reading a specific text type thus does not seem to
have any immediately measurable effects on RMET or Yoni performance.
It is important to clarify here what this outcome means given the Bayesian
confirmative approach that we apply. Although on the face of it this finding represents a null
effect, our approach allows us to quantify exactly what level of support there is in the data for
this ‘null’ hypothesis. That is, there is an 83% chance that preferring this hypothesis over the
other evaluated hypotheses is the right choice for the RMET analysis (and an 81% chance for
the Yoni analysis). It is thus not the case that we can only conclude that we cannot reject the
null hypothesis, but that we don’t really know whether the participants in the various text
30
conditions differ from each other or not (as would be the case in classical null hypothesis
testing). Instead, the results of this analysis show very clear support in favour of the
assumption that the participants in the various text conditions do not differ from each other in
their ToM scores.
Note, though, that the descriptive statistics presented in Table 2 suggest that there are
differences between groups regarding the exposure to literary narrative fiction over the
lifetime and, by consequence, in overall exposure to narrative fiction (i.e., ART literary
scores, and thus also ART overall scores, are lower for the extensive manipulation and the
non-fiction text groups than for the original and the limited manipulation text groups). As
these differences may have influenced the results regarding the direct effects of reading
different text types on ToM, post-hoc analyses were run to investigate this issue. These
analyses are identical to the ones reported above, but they also controlled for the ART literary
scoresvii. These analyses yielded essentially the same results as those presented above. That
is, the specific PMP values of the models change if the ART scores are added to the analyses,
but hypothesis 4-DE is still clearly the most supported by the data (see Table 5). There is thus
no interaction between the text condition that participants were placed in and scores on the
ART.
31
Table 5
Direct effects of text condition on ToM (RMET and Yoni) controlling for ART literary
1-DE: More literary better ToM 0.03 0.12
2-DE: Original text better ToM 0.07 0.13
3-DE: Any literature better ToM 0.03 0.08
4-DE: No differences 0.63 0.58
Cumulative Effects: Narrative Fiction and ToM over the Lifetime
Table 6 shows the results for the first three hypotheses regarding effects of lifetime
exposure to general narrative fiction on ToM.
Table 6
Cumulative effects of general narrative fiction exposure on ToM (RMET and Yoni)
1-CE: + relationship 0.11 0.30
2-CE: - relationship 0.07 0.03
3-CE: No relationship 0.73 0.50
These results demonstrate that for both ToM tasks, there is the most support in the
data for the claim that there is no relationship between lifetime exposure to general narrative
fiction and ToM. This is clearest for the RMET, given that there is a 73% chance that
32
preferring this hypothesis is the right choice. The results are somewhat less clear-cut for the
Yoni task. Although there is a 30% chance that the hypothesis that there is a positive
relationship between the two is correct, the hypothesis that is most supported (PMP of .50) is
the one that states that there is no relationship between ToM and exposure to general
narrative fiction. General exposure to narrative fiction thus does not seem to be correlated
positively with ToM.
Table 7 provides the analyses regarding the relation between exposure to literary and
popular narrative fiction and ToM.
Table 7
Cumulative effects of literary/popular fiction exposure on ToM (RMET and Yoni)
4-CE: Literary > 0 AND literary > popular 0.15 0.55
5-CE: Literary = popular 0.34 0.15
6-CE: Literary = 0 0.42 0.09
Interestingly, the results suggest that there is a difference between the two tasks that
assess ToM in their relation to lifetime exposure to literary and popular narrative fiction.
Although for the RMET there is the most support in the data for Hypothesis 6-CE, according
to which there is no relationship between exposure to literary narrative fiction and ToM, for
the Yoni task Hypothesis 4-CE was most supported. That is, the data suggest that there is a
positive relationship between exposure to literary narrative fiction and Yoni task performance
and that this relationship is stronger than the relationship between exposure to popular
narrative fiction and Yoni task performance. These results are underscored by the correlations
33
(see Table 3): Although the correlation between the RMET and the recognition of both
literary and popular authors is close to zero, the Yoni task shows a dissociation between the
two categories. Whereas Yoni task performance is not correlated with recognition of popular
authors, this correlation is present (if relatively small, r = .18) for the recognition of literary
authors. The more participants recognise literary authors, the higher their scores on the Yoni
task, but higher recognition of popular authors does not have this effect on Yoni scores (and
note that this holds even though there is a high correlation, r = .71, between recognition of
popular and literary authors).
This dissociation between the RMET and the Yoni task with respect to the lifetime
exposure hypotheses was surprising. On the basis of previous studies it was expected that
scores on the ART in general and the literary ART in particular would be positively related to
both measures of ToM. However, the current results suggest that only performance on the
Yoni task is related to narrative fiction exposure and that this relationship is specific to
literary narrative fiction. Given the fact that the Yoni task is comprised of various
subcomponents, the following post-hoc analyses were aimed at investigating which of these
subcomponents is most closely related to exposure to literary narrative fiction. As detailed in
the Procedure section, the Yoni task assesses both understanding of cognitive and affective
mental states at both the first and the second order level. In Table 8, the correlations between
each of these subcomponents and the scores on the literary ART are provided (and, for
completeness’ sake, correlations with the popular ART score have also been added).
34
Table 8
Correlations ART Literary (and Popular) with Yoni subcomponents
These results demonstrate that it is the cognitive subcomponent of the Yoni task that
is most strongly correlated with ART literary scores. The affective subcomponents all have
correlations close to zero with the ART literary score, but the overall cognitive score (which
consists of the total score of the cognitive items measured at the first and second order level)
reaches the highest correlation (r = .21). This finding for the Yoni task is also in accord with
the classification of the RMET as a measure of affective ToM. Evidently, the ability to
attribute affective mental states to others (as indexed by the affective subcomponents of the
Yoni task and the RMET) does not seem to be related to lifetime exposure to literary
narrative fiction, but understanding of others’ cognitive mental states (as indexed by the
cognitive subcomponents of the Yoni task) is related to exposure to this type of fiction. Note,
though, that this finding only holds for lifetime exposure to narrative fiction in relation to
Yoni task performance. Reanalysis of the intervention data to consider only performance on
the cognitive subcomponent of the Yoni task as a function of text condition yielded the same
results as the analysis that considered full Yoni scores as a dependent variable (i.e., there was
still by far the most support in the data for the hypothesis that there was no difference
between text type groups in their performance on the Yoni task).
Discussion
35
Direct Effects of Exposure to (Literary) Narrative Fiction on ToM
Although some studies have not found reading literary narrative fiction to be an
immediate ToM stimulant (Dijkstra et al., 2015; Panero et al., 2016), various other studies
have suggested that there is a positive relation between the two (Kidd & Castano, 2013; Kidd,
et al., 2016; Pino & Mazza, 2016). It is thus currently unclear whether reading literary
narrative fiction directly enhances ToM or not. Furthermore, given that the studies that have
found a direct effect of reading literature on ToM have generally selected works of literary
narrative fiction on the basis of critical acclaim, it is currently also unclear what the ‘active
ingredient’ (or ingredients) for ToM enhancement in literary narrative fiction might be. The
present study thus considered both whether evidence for a direct link between reading
literature and ToM would be found and whether the presence of literary ‘gaps’, as a marker
of literariness, would be relevant in explaining a potential positive association between the
two domains.
Results of the current study demonstrated that brief exposure to narrative fiction did
not directly enhance participants’ ToM scores, regardless of the text’s literariness. All
evaluated hypotheses that assumed that narrative fiction would be an effective ToM stimulant
(be it narrative fiction in a general sense or dependent on its level of literariness) failed to
gain any substantial support in the data. However, the data of the current study supported the
hypothesis that there were no differences in the ToM scores of the various groups of
participants. Given that a Bayesian confirmative approach was applied to the dataset, this
outcome should not be interpreted as a classical null effect. That is, it is not the case that we
simply cannot reject the null hypothesis, instead, the data show very clearly that there is
evidence in favour of the hypothesis that the groups do not differ. The current study thus did
not find evidence for direct effects of exposure to (literary) narrative fiction on ToM ability
36
and the presence of literary ‘gaps’ was not found to play a role in the relationship between
literary narrative fiction and ToM.
However, a number of caveats regarding this conclusion should be kept in mind. In
the first place, the current study only used one text (and two adapted versions of this text) to
act as a ToM stimulant. Regardless of the particular outcome, caution is needed in
generalising findings obtained from one text to (literary) narrative fiction in general.
Although this particular text was chosen because it was part of a set of texts that was found to
be effective in Kidd and Castano (2013), it is possible that this particular text in itself does
not lead to any ToM gains (in this respect it should also be noted that Kidd & Castano, 2016,
found this text to have a low reflective marker score, which suggests that it may not be a
prototypical example of literary narrative fiction). Furthermore, it is also possible that the
specific literary feature that characterised the text that we used (i.e., literary gaps regarding
the source of specific utterances and the social situation of characters) may not be an
important feature in stimulating ToM. Although the presence of gaps is a characteristic
feature of literature and, from a theoretical point of view, this feature can plausibly be
connected to ToM enhancement (Koopman & Hakemulder, 2015), it could be that this
feature (either the presence of gaps in general or as specifically operationalised in the current
study) is simply not directly relevant in providing readers with ToM gains.
Alternatively, it is possible that the literary text (and potentially one or both of its
adaptations) did have an immediate effect on the reader’s ToM, but it was too subtle to pick
up in our measures or with our sample size (although our measures and sample size were
similar to those used in Kidd & Castano, 2013). Studies that employ more texts and that
manipulate other types of gap filling may thus provide a different picture of the relationship
between literary gap-filling and direct ToM gains.
37
Future work might examine both text and reader variables to consider what factors
determine whether a particular work of narrative fiction has a direct effect on the reader’s
ToM. On the reader side, it may be that some readers are more likely to profit from exposure
to literary narrative fiction in terms of ToM gains. For instance, it may be the case that
readers who are more prone to being transported into fictional worlds (i.e., readers with high
levels of transportability) profit more from reading about complex others’ mental states (cf.
Bal & Veltkamp, 2013). On the text side, a deeper investigation into the function of gaps may
demonstrate that they are relevant after all. As argued above, literary gaps may occur in
various forms, so perhaps different gap forms may be more powerful than the ones
manipulated in the present study. Moreover, it is clear that there are many candidates for
further investigation, such as: Foregrounding, polyvalence (or multi-interpretability),
ambiguity, polyphony (the co-presence of different narratorial ‘voices’), characters’
roundness (e.g., having various contrasting traits), level of access to characters’ inner worlds,
the development of a character’s personality etcetera. These may all be taken into
consideration for future manipulation. Of course, none of these text qualities are unique to
literature, but they are often assumed to constitute the degree of ‘literariness’ and seem to be
dominantly found in literature, and less so in popular fiction. More importantly, it seems
plausible that they contribute to the development of readers’ ToM – if not after exposure to
one short story, then maybe cumulatively, after years of literary reading.
Cumulative Effects of Exposure to Narrative Fiction over the Lifetime on ToM
Our results suggest that there is no relationship between lifetime exposure to general
narrative fiction and ToM. Furthermore, the results of the analyses regarding literary
narrative fiction exposure across the lifetime and ToM were somewhat mixed. Performance
on the RMET (as a measure of affective ToM) was not related to exposure to literary
38
narrative fiction (in contrast to the findings reported in Kidd & Castano, 2016). However,
there was a positive relationship between Yoni task scores (as a measure of both cognitive
and affective ToM) and lifetime literary narrative fiction exposure and this relationship was
found to be stronger than that between lifetime popular narrative fiction exposure and Yoni
scores. Further analyses of the Yoni task results showed that the strongest positive correlation
was present between performance on the cognitive component of the Yoni task and literary
narrative fiction. These results thus suggest that lifetime exposure to literary narrative fiction
is associated with higher levels of cognitive ToM (understanding what other people are
thinking about), but not with affective ToM (understanding what other people are feeling). It
is relevant to note here that if scores on the literary ART measure simply indicated more
exposure to narrative fiction in general instead of being a specific indicator of exposure to
literary narrative fiction, we would expect both the general ART scores and the literary ART
scores to be related to ToM measures. As this is not the case (despite the high correlation
between literary and popular ART scores), this suggests that there is something specific to
exposure to literary narrative fiction that is responsible for this observed relationship.
Although the results initially seem to go against findings from previous studies that
suggest that there is a relationship between ToM and general narrative fiction lifetime
exposure (Djikic et al., 2013; Mar et al., 2006, 2009; Panero et al., 2016), the differences
between the current study and previous work may not be profound. Potentially, the positive
relation between general narrative fiction exposure and ToM that has been found in previous
studies actually boils down to a positive correlation between exposure to literary narrative
fiction and (cognitive) ToM. Given that previous studies on this topic have generally not
divided narrative fiction exposure into literary and popular categoriesviii (but see Kidd &
Castano, 2016, for an exception), they may have primarily been measuring exposure to
literary narrative fiction in their general narrative fiction exposure measure. The current study
39
may then not have found a positive correlation between general narrative fiction and ToM
because equal numbers of literary and popular authors were present in the ART. Overall
recognition scores could thus only be driven by literary narrative fiction exposure for
maximally fifty percent, whereas this percentage may be (much) higher in other studies.
Future research is needed to determine whether a distinction between literary and popular
narrative fiction in exposure measures will consistently demonstrate stronger relations
between literary narrative fiction and ToM as compared to popular narrative fiction.
Although we had expected lifetime exposure to literary narrative fiction to be
positively related to ToM, we did not predict that this result would only hold for the Yoni
task and not for the RMET, especially given that previous studies have found evidence for a
positive relationship between narrative fiction exposure and RMET performance (e.g., Djikic
et al., 2013; Panero et al., 2016) and even specifically between literary narrative fiction
exposure and RMET performance (Kidd & Castano, 2016). However, only for the Yoni task
did we find that the ToM scores were positively correlated with literary narrative fiction
exposure (and more strongly so than for popular narrative fiction exposure). Further analyses
demonstrated that this relationship was primarily driven by the cognitive subcomponent of
the Yoni task; correlations between literary narrative fiction exposure and the affective
subcomponent were close to zero. There is thus an internal consistency in the findings: Those
ToM measures that assess cognitive ToM (i.e., the cognitive subcomponents of the Yoni task)
are positively correlated with literary narrative fiction exposure, whereas those that assess
affective ToM (the affective subcomponents of the Yoni task and the RMET) are not. Our
data thus suggests that exposure to literary narrative fiction is related to knowledge about
others’ thoughts and not so much to knowledge about others’ emotions. Interestingly,
although our results contrast with those of Kidd and Castano (2016) who do find a positive
correlation between exposure to literary narrative fiction and RMET performance, the small
40
number of other studies that have also made a distinction between cognitive vs. affective
ToM in relation to (literary) narrative fiction exposure find similar results to the current
study. For instance, Pino and Mazza (2016) found that exposure to literary narrative fiction
positively affects mentalising (the ability to understand others’ mental states), but not
emotion sharing abilities (the capacity to emotionally resonate with others’ feelings). The
findings reported by Stansfield and Bunce (2014) also closely parallel the current results, as
they did find a positive correlation between exposure to narrative fiction and cognitive ToM
(as assessed using a questionnaire measuring perspective taking tendency), but did not find
evidence for a correlation between narrative fiction exposure and scores on the RMET.
These results thus suggest that there might be a special relationship between exposure
to (literary) narrative fiction and cognitive ToM. Indeed, it is known from the literature that
cognitive and affective ToM are dissociable and are sub-served by different neural networks
(Kalbe et al., 2010; Shamay-Tsoory & Aharon-Peretz, 2007; Shamay-Tsoory et al., 2007). In
principle, it is thus possible that reading literary narrative fiction is specifically stimulating
for readers’ ability to understand the thoughts of others, whilst it may not have any
demonstrable effects on their ability to understand others’ emotions. It is characteristic of
literature that its plotlines and character interactions are complex at the psychological level
(Gerrig & Rapp, 2004; Koopman & Hakemulder, 2015; Van Peer, 1986). Potentially, this
complexity pertains more to being able to fathom others’ thoughts (at both the author and
character level) than their emotions and thus exposure to literary narrative fiction is most
effective at stimulating our cognitive ToM, whilst affective ToM is not enhanced.
However, it should be noted that given the correlational nature of the relationship
between lifetime exposure to literary narrative fiction and cognitive ToM, it is not possible to
draw firm conclusions regarding the causal direction of this link. Exposure to literary
narrative fiction may have a beneficial effect on readers’ ability to understand others’
41
thoughts, but it is also possible that those readers who have a more advanced cognitive ToM
are those who feel particularly drawn to literary narrative fiction. Additional intervention
studies that assess the effects of literary narrative fiction specifically on cognitive vs.
affective ToM are needed in order for stronger conclusions to be drawn regarding the
relationship between exposure to literary narrative fiction and cognitive ToM.
In summary, although the results of the intervention component of the current study
suggest that there are no directly measurable ToM gains to be had from a one-off encounter
with a piece of literary narrative fiction (or at least not the short piece of literary narrative
fiction used in this study), there do seem to be interesting links between lifetime exposure to
literary narrative fiction and cognitive ToM. One short story by Chekhov might thus not get
you very far as mental state understanding goes, but, chances are, the book smarts you gain
from reading his collected works will spill over to your real world street smarts.
42
References
Adrian, J. E., Clemente, R. A., Villanueva, L., & Rieffe, C. (2005). Parent–child picture-book
reading, mothers' mental state language and children's theory of mind. Journal of Child
Language, 32(3), 673-686. http://dx.doi.org/10.1017/S0305000905006963
Aram, D., & Aviram, S. (2009). Mothers' storybook reading and kindergartners'
socioemotional and literacy development. Reading Psychology, 30(2), 175-194.
http://dx.doi.org/10.1080/02702710802275348
Bal, P. M., & Veltkamp, M. (2013). How does fiction reading influence empathy? An
experimental investigation on the role of emotional transportation. PloS one, 8(1),
e55341. http://dx.doi.org/10.1371/journal.pone.0055341
Barnes, J. L. (2012). Fiction, imagination, and social cognition: Insights from autism. Poetics,
40(4), 299-316. http://dx.doi.org/10.1016/j.poetic.2012.05.001
Baron‐Cohen, S., Wheelwright, S., Hill, J., Raste, Y., & Plumb, I. (2001). The “Reading the
Mind in the Eyes” test revised version: A study with normal adults, and adults with
Asperger syndrome or high‐functioning autism. Journal of Child Psychology and
Psychiatry, 42(2), 241-251. http://dx.doi.org/10.1111/1469-7610.00715
Black, J. E., & Barnes, J. L. (2015). The effects of reading material on social and non-social
cognition. Poetics, 52, 32-43. http://dx.doi.org/10.1016/j.poetic.2015.07.001
Chekhov, A. (1884/1978). A chameleon. In R. E. Matlow (Ed.), Anton Chekhov’s short
stories (pp. 1-3). New York, NY: Norton.
Chekhov, A. (1884/2014). De kameleon [The chameleon]. In Verzamelde werken 1 verhalen
1880-1885 (pp. 285-288). Amsterdam, The Netherlands: Van Oorschot.
Dijkstra, K., Verkoeijen, P. P. J. L., van Kuijk, I., Yee Chow, S., Bakker, A. & Zwaan, R.A.
(2015). Leidt het lezen van literaire fictie tot meer empathie? Een replicatiestudie [Does
43
reading literature result in higher empathy? A replication study]. De Psycholoog, 50(10),
10-21.
Djikic, M., Oatley, K., & Moldoveanu, M. C. (2013). Reading other minds: Effects of
literature on empathy. Scientific Study of Literature, 3(1), 28-47.
http://dx.doi.org/10.1075/ssol.3.1.06dji
Eugenides, J. (1993). The virgin suicides. New York, NY: Picador.
Fong, K., Mullin, J. B., & Mar, R. A. (2013). What you read matters: The role of fiction
genre in predicting interpersonal sensitivity. Psychology of Aesthetics, Creativity, and the
Arts, 7(4), 370-376. http://dx.doi.org/10.1037/a0034084
Gamannossi, B. A., & Pinto, G. (2014). Theory of mind and language of mind in narratives:
Developmental trends from kindergarten to primary school. First Language, 34(3), 262–
272. http://dx.doi.org/10.1177/0142723714535875
Gerrig, R. & Rapp, D. (2004). Psychological processes underlying literary impact. Poetics
Today, 25(2), 265–281. http://dx.doi.org/10.1215/03335372-25-2-265
Hakemulder, J. (2000). The moral laboratory: Experiments examining the effects of reading
literature on social perception and moral self-concept. Amsterdam, The Netherlands:
John Benjamins Publishing.
Harlow, L. L., Mulaik, S. A., & Steiger, J. H. (1997). What if there were no significance
tests? Mahwah, NJ: Erlbaum.
Hoijtink, H. (2012). Informative hypotheses: Theory and practice for behavioral and social
scientists. Boca Raton, FL: CRC.
Iser, W. (1975). Die Appellstruktur der Texte. In R. Warning (Ed.), Rezeptionsästhetik:
Theorie und Praxis (pp. 228-252). Munich, Germany: Fink.
Iser, W. (1978). The act of reading: A theory of aesthetic response. Baltimore, MD: Johns
Hopkins University Press.
44
Iser, W. (1988). The reading process: A phenomenological approach. In D. Lodge (Ed.),
Modern criticism and thought: A reader (pp. 279- 299). London, United Kingdom:
Longman.
Kalbe, E., Schlegel, M., Sack, A. T., Nowak, D. A., Dafotakis, M., Bangard, C., ... & Kessler,
J. (2010). Dissociating cognitive from affective theory of mind: A TMS study. Cortex,
46(6), 769-780. http://dx.doi.org/10.1016/j.cortex.2009.07.010
Keen, S. (2007). Empathy and the novel. New York, NY: Oxford University Press.
Kidd, D. C., & Castano, E. (2013). Reading literary fiction improves theory of mind. Science,
342(6156), 377-380. http://dx.doi.org/10.1126/science.1239918
Kidd, D., & Castano, E. (2016). Different stories: How levels of familiarity with literary and
genre fiction relate to mentalizing. Psychology of Aesthetics, Creativity, and the
Arts, Advance online publication. http://dx.doi.org/10.1037/aca0000069
Kidd, D. C., & Castano, E. (2017). Panero et al. (2016): Failure to replicate methods caused
the failure to replicate results. Journal of Personality and Social Psychology, 112(3), e1–
e4. http://dx.doi.org/10.1037/pspa0000072
Kidd, D., Ongis, M., & Castano, E. (2016). On literary fiction and its effects on theory of
mind. Scientific Study of Literature, 6(1), 42-58. http://dx.doi.org/10.1075/ssol.6.1.04kid
Koopman, E. M. E. (2015a). Empathic reactions after reading: The role of genre, personal
factors and affective responses. Poetics, 50, 62-79.
http://dx.doi.org/10.1016/j.poetic.2015.02.008
Koopman, E. M. E. (2015b). How texts about suffering trigger reflection: Genre, personal
factors and affective responses. Psychology of Aesthetics, Creativity, and the Arts, 9(4),
430–441. http://dx.doi.org/10.1037/aca0000006
Koopman, E. M. E. (2016a). Reading suffering. Rotterdam, The Netherlands: ERMeCC,
Erasmus Research Center for Media, Communication and Culture
45
Koopman, E. M. E. (2016b). Effects of “literariness” on emotions and on empathy and
reflection after reading. Psychology of Aesthetics, Creativity, and the Arts, 10(1), 82-98.
http://dx.doi.org/10.1037/aca0000041
Koopman, E. M. E., & Hakemulder, F. (2015). Effects of literature on empathy and self-
reflection: A theoretical-empirical framework. Journal of Literary Theory, 9(1), 79-111.
http://dx.doi.org/10.1515/jlt-2015-0005
Kluytmans, A., van de Schoot, R., Mulder, J., & Hoijtink, H. (2012). Illustrating Bayesian
evaluation of informative hypotheses for regression models. Frontiers in Psychology,
3(2), 1-11. http://dx.doi.org/10.3389/fpsyg.2012.00002
Mar, R. A., & Oatley, K. (2008). The function of fiction is the abstraction and simulation of
social experience. Perspectives on Psychological Science, 3(3), 173-192.
http://dx.doi.org/10.1111/j.1745-6924.2008.00073.x
Mar, R. A., Oatley, K., Hirsh, J., dela Paz, J., & Peterson, J. B. (2006). Bookworms versus
nerds: Exposure to fiction versus non-fiction, divergent associations with social ability,
and the simulation of fictional social worlds. Journal of Research in Personality, 40(5),
694-712. http://dx.doi.org/10.1016/j.jrp.2005.08.002
Mar, R. A., Oatley, K., & Peterson, J. B. (2009). Exploring the link between reading fiction
and empathy: Ruling out individual differences and examining outcomes.
Communications, 34(4), 407-428. http://dx.doi.org/10.1515/COMM.2009.025
Mol, S. E., & Bus, A. G. (2011). To read or not to read: A meta-analysis of print exposure
from infancy to early adulthood. Psychological Bulletin, 137(2), 267-296.
http://dx.doi.org/10.1037/a0021890
Moore, M., & Gordon, P. C. (2015). Reading ability and print exposure: Item response theory
analysis of the author recognition test. Behavior Research Methods, 47(4), 1095–1109.
http://dx.doi.org/10.3758/s13428-014-0534-3
46
Mulder, J., Hoijtink, H. & Klugkist, I. (2009). Equality and inequality constrained
multivariate linear models: Model selection using constrained posterior priors. Journal of
Statistical Planning and Inference, 140(4), 887-906.
http://dx.doi.org/10.1016/j.jspi.2009.09.022
Mulder, J., Hoijtink, H., & de Leeuw, C. (2012). BIEMS: A Fortran 90 program for
calculating Bayes factors for inequality and equality constrained models. Journal of
Statistical Software, 46(2), 1-39. http://dx.doi.org/10.18637/jss.v046.i02
Mulder, J., Klugkist, I. G., Schoot, A. G. J. van de, Meeus, W. H. J., Selfhout, M. H. W. &
Hoijtink, H. J. A. (2009). Informative hypotheses for repeated measurements: A
Bayesian approach. Journal of Mathematical Psychology, 53(6), 530-546.
http://dx.doi.org/10.1016/j.jmp.2009.09.003
Mumper, M. L. & Gerrig, R. J. (2016). Leisure reading and social cognition: A meta-analysis.
Psychology of Aesthetics, Creativity, and the Arts, 11(1), 109-120.
http://dx.doi.org/10.1037/aca0000089
Oatley, K. (1999a). Why fiction may be twice as true as fact: Fiction as cognitive and
emotional simulation. Review of General Psychology, 3(2), 101.
http://dx.doi.org/10.1037/1089-2680.3.2.101
Oatley, K. (1999b). Meetings of minds: Dialogue, sympathy, and identification, in reading
fiction. Poetics, 26(5-6), 439-454. http://dx.doi.org/10.1016/S0304-422X(99)00011-X
Oatley, K., & Mar, R. A. (2005). Evolutionary pre-adaptation and the idea of character in
fiction. Journal of Cultural and Evolutionary Psychology, 3(2), 179-194.
http://dx.doi.org/10.1556/JCEP.3.2005.2.5
Panero, M. E., Weisberg, D. S., Black, J., Goldstein, T. R., Barnes, J. L., Brownell, H., &
Winner, E. (2016). Does reading a single passage of literary fiction really improve theory
47
of mind? An attempt at replication. Journal of Personality and Social Psychology,
111(5), e46-e54. http://dx.doi.org/10.1037/pspa0000064
Peskin, J., & Astington, J. W. (2004). The effects of adding metacognitive language to story
texts. Cognitive Development, 19(2), 253-273.
http://dx.doi.org/10.1016/j.cogdev.2004.01.003
Pino, M. C., & Mazza, M. (2016). The use of “literary fiction” to promote mentalizing
ability. PloS one, 11(8), e0160254. http://dx.doi.org/10.1371/journal.pone.0160254
Ratner, N. K., & Olver, R. R. (1998). Reading a tale of deception, learning a theory of mind?
Early Childhood Research Quarterly, 13(2), 219-239. http://dx.doi.org/10.1016/S0885-
2006(99)80036-2
Rosnow, R. L., & Rosenthal, R. (1989). Statistical procedures and the justification of
knowledge in psychological science. American Psychologist, 44(10), 1276-1284.
http://dx.doi.org/10.1037/0003-066X.44.10.1276
Shamay-Tsoory, S. G., & Aharon-Peretz, J. (2007). Dissociable prefrontal networks for
cognitive and affective theory of mind: A lesion study. Neuropsychologia, 45(13), 3054-
3067. http://dx.doi.org/10.1016/j.neuropsychologia.2007.05.021
Shamay-Tsoory, S. G., Shur, S., Barcai-Goodman, L., Medlovich, S., Harari, H., &
Levkovitz, Y. (2007). Dissociation of cognitive from affective components of theory of
mind in schizophrenia. Psychiatry Research, 149(1), 11-23.
http://dx.doi.org/10.1016/j.psychres.2005.10.018
Shamay-Tsoory, S. G., Tibi-Elhanany, Y., & Aharon-Peretz, J. (2007). The green-eyed
monster and malicious joy: The neuroanatomical bases of envy and gloating
(schadenfreude). Brain, 130(6), 1663-1678. http://dx.doi.org/10.1093/brain/awm093
Stanovich, K. E., & West, R. F. (1989). Exposure to print and orthographic processing.
Reading Research Quarterly, 24(4), 402–433. http://dx.doi.org/10.2307/747605
48
Stansfield, J. & Bunce, L. (2014). The relationship between empathy and reading fiction:
Separate roles for cognitive and affective components. Journal of European Psychology
Students, 5(3), 9-18. http://dx.doi.org/10.5334/jeps.ca
Sugiyama, M. S. (2001). Food, foragers, and folklore: The role of narrative in human
subsistence. Evolution and Human Behavior, 22(4), 221-240.
http://dx.doi.org/10.1016/S1090-5138(01)00063-0
van Beem, J. (2015). Kleur bekennen [Follow suit]. National Geographic, 9, 118-130.
van de Schoot, R., Kaplan, D., Denissen, J., Asendorpf, J. B., Neyer, F. J., & Aken, M. A.
(2014). A gentle introduction to Bayesian analysis: Applications to developmental
research. Child Development, 85(3), 842-860. http://dx.doi.org/10.1111/cdev.12169
van de Schoot, R., Mulder, J., Hoijtink, H., Van Aken, M. A., Semon Dubas, J., Orobio de
Castro, B., ... & Romeijn, J. W. (2011). An introduction to Bayesian model selection for
evaluating informative hypotheses. European Journal of Developmental Psychology,
8(6), 713-729. http://dx.doi.org/10.1080/17405629.2011.621799
Van Peer, W. (1986). Pulp and purpose. Stylistic analysis as an aid to a theory of texts. In T.
D’Haen (Ed.), Linguistics and the study of literature (pp. 268–286). Amsterdam, The
Netherlands: Rodopi.
Van Peer, W. (1989). Quantitative studies of literature. A critique and an outlook. Computers
and the Humanities, 23(4), 301-307. http://dx.doi.org/10.1007/BF02176635
West, R. F., Stanovich, K. E., & Mitchell, H. R. (1993). Reading in the real world and its
correlates. Reading Research Quarterly, 28(1), 34–50. http://dx.doi.org/10.2307/747815
Zunshine, L. (2006). Why we read fiction: Theory of mind and the novel. Columbus, OH:
Ohio State University Press.
49
Appendix: Texts
A Chameleon, original version (English version provided here, Dutch translation used)
The police superintendent Otsjoemelov is walking across the market square wearing a new
overcoat and carrying a parcel under his arm. A red-haired policeman strides after him with a
sieve filled to the rim with confiscated gooseberries. There is silence all around... Not a soul
in the square... The open doors of the shops and taverns appear disconsolately, like hungry
mouths into the wide world; there is not even a beggar near them.
“So you bite, you damned brute?” Otsjoemelov hears suddenly. “Lads, don’t let it go! Biting
is prohibited nowadays! Hold it! Ah... ah!”
There is the sound of a dog yelping. Otsjoemelov looks in the direction of the sound
and sees a dog, hopping on three legs and looking about it, run out of merchant Pitsjoegin’s
timber-yard. A man in a starched cotton shirt, with his waistcoat unbuttoned, is chasing it.
The man runs after the animal, and throwing his body forward falls down and seizes the dog
by its hind legs. Once more there is a yelping and a sound of “Don’t let go!” Sleepy
countenances are protruded from the shops, and soon a crowd, as if out of nowhere, is
gathered round the timber-yard.
“It looks like a row, your honour...” says the policeman.
Otsjoemelov makes half a turn to the left and strides towards the crowd. He sees the
aforementioned man in the unbuttoned waistcoat standing close by the gate of the timber-
yard, holding his right hand in the air and displaying a bleeding finger to the crowd. On his
half-drunken face there is plainly written: “I’ll pay you back, you rogue!” and indeed the very
finger has the look of a flag of victory. In this man Otsjoemelov recognises Chrjoekin, the
goldsmith. The culprit who has caused the sensation, a white borzoy puppy with a sharp
muzzle and a yellow patch on its back, is sitting on the ground with its fore-paws outstretched
50
in the middle of the crowd, trembling all over. There is an expression of misery and terror in
its tearful eyes.
“What’s it all about?” Otsjoemelov inquires, pushing his way through the crowd.
“What are you here for? Why are you waiving your finger?.. Who shouted?”
“I was walking along here, not interfering with anyone, your honour...” Chrjoekin
begins, coughing into his fist. “Because of firewood, with Mitri Mitritsj, this low brute for no
rhyme or reason bit my finger... You must excuse me, I am a working man... Mine is fine
work. I must have damages, for I shan’t be able to use this finger for a week, maybe... It’s not
even the law, your honour, that one should put up with it from a beast... If everyone is going
to be bitten, life on earth won’t be worth living...”
“Hm!.. Good...” says Otsjoemelov sternly, coughing and raising his eyebrows.
“Good... Whose dog is it? I won’t let this pass! I’ll teach you to let your dogs run all over the
place! It’s time these gentry were looked after, if they won’t obey the regulations! When he’s
fined, the blackguard, I’ll teach him what it means to keep dogs and such stray cattle! I’ll
give him a lesson!.. Jeldyrin,” says the superintendent, addressing the policeman, “find out
whose dog this is and draw up a report! And the dog must be destroyed. Without delay! It’s
sure to be mad... Whose dog is it? I ask!”
“I fancy it’s General Zjigalov’s,” says someone in the crowd.
“General Zjigalov’s? Hm!.. Help me off with my coat, Jeldyrin... it’s frightfully hot!
It must be a sign of rain... There’s one thing I can’t make out, how it came to bite you?”
Otsjoemelov asks Chrjoekin. “Surely he couldn’t reach your finger. He’s a little dog, and you
are a great hulking fellow! You must have scratched your finger with a nail, and then the idea
struck you to get damages for it. We all know... your sort! I know you devils!”
“He put a cigarette in its face, your honour, for a joke, but it had the sense to snap at
him... He is a nonsensical fellow, your honour!”
51
“That’s a lie, Squinteye! You didn’t see, so why tell lies about it? His honour is a wise
gentleman, and will see who is telling lies and who is telling the truth, as in God’s sight...
And if I am lying let the court decide. It’s written in the law... We are all equal nowadays.
My own brother is in the gendarmes... if you want to know...”
“Shut your mouth!”
“No, that’s not the General’s dog,” says the policeman, with profound conviction,
“the General hasn’t got ones like that. His are mostly setters.”
“Do you know that for a fact?”
“Yes, your honour.”
“I know it, too. The General has valuable dogs, thoroughbred, and this is goodness
knows what! No coat, no shape... a low creature... And to keep a dog like that?!... Where’s
the sense of it. If a dog like that were to turn up in Petersburg or Moscow, do you know what
would happen? They would not worry about the law, but at once – away with it! You’ve been
injured, Chrjoekin, and we can’t let the matter drop... We must give them a lesson! It is high
time...”
“Yet maybe it is the General’s...” says the policeman, thinking aloud. “It’s not written
on its face... I saw one like it the other day in his yard.”
“It’s the General’s, that’s certain!” says a voice in the crowd.
“Hm!.. help me on with my overcoat, Jeldyrin, my lad... The wind’s getting up... I am
shivering... Take the dog to the General, and inquire there. Say I found it and sent you... And
tell them not to let it out into the street... It may be a valuable dog, and if every swine goes
sticking a cigar to its nose, it will soon be ruined. A dog is a delicate animal... And you put
your hand down, you blockhead! It’s no use your displaying your fool of a finger! It’s your
own fault!....”
52
“Here comes the General’s cook, ask him... Hi, Prochor! Come here, my dear man!
Do you see that dog?... Is it yours?”
“Are you kidding? We have never had one like that!”
“There’s no need to waste time asking,” says Otsjoemelov. “It’s a stray dog! There’s
no need to waste time talking about it... If I say it’s a stray dog, a stray dog it is... It must be
destroyed, that’s all about it.”
“It is not our dog,” Prochor goes on. “It belongs to the General’s brother, who arrived
the other day. Our master does not care for hounds. But his honour is fond of them...”
“You don’t say his brother has arrived? Vladimir Ivanytsj?” inquires Otsjoemelov,
and his whole face beams with an ecstatic smile. “Well, I never! And I didn’t know! Has he
come on a visit?
“Yes...”
“Well, I never... He was missing his brother... And there I didn’t know! So this is his
honour’s dog? Delighted to hear it... Take it... Good dog... A lively creature. . . . Snapped at
this fellow’s finger! Ha-ha-ha... Come on, why are you shivering? Grrr... Grrr... The rogue’s
angry... a nice little pup.”
Prochor calls the dog and walks away from the timber-yard with it. The crowd laughs
at Chrjoekin.
“I’ll get you!” Otsjoemelov threatens him, and wrapping himself in his greatcoat, goes
on his way across the square.
53
A Chameleon, limited manipulation version (English version provided here, Dutch
translation used; changes marked in boldface)
The police superintendent Otsjoemelov is walking across the market square wearing a new
overcoat and carrying a parcel under his arm. A The red-haired policeman Jeldyrin strides
after him with a sieve filled to the rim with confiscated gooseberries. There is silence all
around... Not a soul in the square... The open doors of the shops and taverns appear
disconsolately, like hungry mouths into the wide world; there is not even a beggar near them.
“So you bite, you damned brute?” Otsjoemelov hears suddenly. “Lads, don’t let it
that dog go! Biting is prohibited nowadays! Hold it! Ah... ah!”
There is the sound of a dog yelping. Otsjoemelov looks in the direction of the sound
and sees a dog, hopping on three legs and looking about it, run out of merchant Pitsjoegin’s
timber-yard. A man in a starched cotton shirt, with his waistcoat unbuttoned, is chasing it.
The man; it is the goldsmith Chrjoekin. Chrjoekin runs after the animal, and throwing his
body forward falls down and seizes the dog by its hind legs. Once more there is a yelping and
a sound of “Don’t let go!” Sleepy countenances are protruded from the shops, and soon a
crowd, as if out of nowhere, is gathered round the timber-yard.
“It looks like a row, your honour...” says the policeman Jeldyrin to Otsjoemelov.
Otsjoemelov makes half a turn to the left and strides towards the crowd. He sees the
aforementioned man in the unbuttoned waistcoat Chrjoekin standing close by the gate of the
timber-yard, holding his right hand in the air and displaying a bleeding finger to the crowd.
On his half-drunken face there is plainly written: “I’ll pay you back, you rogue!” and indeed
the very finger has the look of a flag of victory. In this man Otsjoemelov recognises
Chrjoekin, the goldsmith. The culprit who has caused the sensation, a white borzoy puppy
with a sharp muzzle and a yellow patch on its back, is sitting on the ground with its fore-paws
54
outstretched in the middle of the crowd, trembling all over. There is an expression of misery
and terror in its tearful eyes.
“What’s it all about?” Otsjoemelov inquires, pushing his way through the crowd.
“What are you here for? Why are you waiving your finger?.. Who shouted?”
“I was walking along here, not interfering with anyone, your honour...” Chrjoekin
begins, coughing into his fist. “Because of firewood, with Mitri Mitritsj, this low brute for no
rhyme or reason bit my finger... You must excuse me, I am a working man... Mine is fine
work. I must have damages, for I shan’t be able to use this finger for a week, maybe... It’s not
even the law, your honour, that one should put up with it from a beast... If everyone is going
to be bitten, life on earth won’t be worth living...”
“Hm!.. Good...” says Otsjoemelov sternly, coughing and raising his eyebrows.
“Good... Whose dog is it? I won’t let this pass! I’ll teach you to let their dogs run all over the
give him a lesson!.. Jeldyrin,” says the superintendent, addressing the policeman, “find out
whose dog this is and draw up a report! And the dog must be destroyed. Without delay! It’s
sure to be mad... Whose dog is it? I ask!”
“I fancy it’s General Zjigalov’s,” says someone in the crowd.
“General Zjigalov’s? Hm!.. Help me off with my coat, Jeldyrin... it’s frightfully hot!
It must be a sign of rain... There’s one thing I can’t make out, how it came to bite you?”
Otsjoemelov asks Chrjoekin. “Surely he couldn’t reach your finger. He’s a little dog, and you
are a great hulking fellow! You must have scratched your finger with a nail, and then the idea
struck you to get damages for it. We all know... your sort! I know you devils!”
“He put a cigarette in its face, your honour, for a joke,” says someone in the crowd.
“But but it had the sense to snap at him... He is a nonsensical fellow, your honour!”
55
“That’s a lie, Squinteye!” Chrjoekin calls out. “You didn’t see, so why tell lies about
it? His honour is a wise gentleman, and will see who is telling lies and who is telling the
truth, as in God’s sight... And if I am lying let the court decide. It’s written in the law... We
are all equal nowadays. My own brother is in the gendarmes... if you want to know...”
“Shut your mouth!” Otsjoemelov orders.
“No, that’s not the General’s dog,” says the policeman Jeldyrin, with profound
conviction, “the General hasn’t got ones like that. His are mostly setters.”
“Do you know that for a fact?” Otsjoemelov asks.
“Yes, your honour.”
“I know it, too.,” says Otsjoemelov. “The General has valuable dogs, thoroughbred,
and this is goodness knows what! No coat, no shape... a low creature... And to keep a dog like
that?!... Where’s the sense of it. If a dog like that were to turn up in Petersburg or Moscow,
do you know what would happen? They would not worry about the law, but at once – away
with it! You’ve been injured, Chrjoekin, and we can’t let the matter drop... We must give
them a lesson! It is high time...”
“Yet maybe it is the General’s...” says the policeman Jeldyrin, thinking aloud. “It’s
not written on its face... I saw one like it the other day in his yard.”
“It’s the General’s, that’s certain! “ says a voice in the crowd.
“Hm!.. help me on with my overcoat, Jeldyrin, my lad...” Otsjoemelov requests
Jeldyrin. “The wind’s getting up... I am shivering... Take the dog to the General, and inquire
there. Say I found it and sent you... And tell them not to let it out into the street... It may be a
valuable dog, and if every swine goes sticking a cigar to its nose, it will soon be ruined. A
dog is a delicate animal... And you Chrjoekin, put your hand down, you blockhead! It’s no
use your displaying your fool of a finger! It’s your own fault!....”
56
“Here comes the General’s cook, ask him...” Jeldyrin suggests. “Hi, Prochor! Come
here, my dear man! Do you see that dog?... Is it yours?”
“Are you kidding? We have never had one like that!” Prochor says.
“There’s no need to waste time asking,” says Otsjoemelov. “It’s a stray dog! There’s
no need to waste time talking about it... If I say it’s a stray dog, a stray dog it is... It must be
destroyed, that’s all about it.”
“It is not our dog,” Prochor goes on. “It belongs to the General’s brother, who arrived
the other day. Our master does not care for hounds. But his honour is fond of them...”
“You don’t say his brother has arrived? Vladimir Ivanytsj?” inquires Otsjoemelov,
and his whole face beams with an ecstatic smile. “Well, I never! And I didn’t know! Has he
come on a visit?
“Yes...” Prochor starts.
“Well, I never... He was missing his brother...” Otsjoemelov continues. “And there I
didn’t know! So this is his honour’s dog? Delighted to hear it... Take it... Good dog... A
lively creature. . . . Snapped at this fellow’s finger! Ha-ha-ha... Come on, why are you
shivering? Grrr... Grrr... The rogue’s angry... a nice little pup.”
Prochor calls the dog and walks away from the timber-yard with it. The crowd laughs
at Chrjoekin.
“I’ll get you!” Otsjoemelov threatens him, and wrapping himself in his greatcoat, goes
on his way across the square.
57
A Chameleon, extensive manipulation version (English version provided here, Dutch
translation used; changes marked in boldface)
The police superintendent Otsjoemelov Jansen is walking across the market square wearing a
new overcoat and carrying a parcel under his arm. A The red-haired policeman De Vries
strides after him with a sieve filled to the rim with confiscated gooseberries. There is silence
all around... Not a soul in the square... The open doors of the shops and taverns appear
disconsolately, like hungry mouths into the wide world; there is not even a beggar near them.
“So you bite, you damned brute?” Otsjoemelov Jansen hears goldsmith Tiggelaar
shout suddenly. “Lads, don’t let it that dog go! Biting is prohibited nowadays! Hold it! Ah...
ah!”
There is the sound of a dog yelping. Otsjoemelov Jansen looks in the direction of the
sound and sees a dog, hopping on three legs and looking about it, run out of merchant
Pitsjoegin’s a timber-yard. A man The goldsmith Tiggelaar in a starched cotton shirt, with
his waistcoat unbuttoned, is chasing it. The man Tiggelaar runs after the animal, and
throwing his body forward falls down and seizes the dog by its hind legs. Once more there is
a yelping and a sound of “Don’t let go!” Sleepy countenances are protruded from the shops,
and soon a crowd, as if out of nowhere, is gathered round the timber-yard.
“It looks like a row, your honour boss!..” says the policeman De Vries.
Otsjoemelov Jansen makes half a turn to the left and strides towards the crowd. He
sees the aforementioned man in the unbuttoned waistcoat Tiggelaar standing close by the
gate of the timber-yard, holding his right hand in the air and displaying a bleeding finger to
the crowd. On his half-drunken face there is plainly written: “I’ll pay you back, you rogue!”
and indeed the very finger has the look of a flag of victory. In this man Otsjoemelov
recognises Chrjoekin, the goldsmith. The culprit who has caused the sensation, a white
borzoy puppy with a sharp muzzle and a yellow patch on its back, is sitting on the ground
58
with its fore-paws outstretched in the middle of the crowd, trembling all over. There is an
expression of misery and terror in its tearful eyes.
“What’s it all about?” Otsjoemelov Jansen inquires, pushing his way through the
crowd. “What are you here for? Why are you waiving your finger?.. Who shouted?”
“I was walking along here, not interfering with anyone, your honour inspector...”
Chrjoekin Tiggelaar begins, coughing into his fist. “Because of firewood, with Mitri Mitritsj,
this This low brute for no rhyme or reason bit my finger...!.. You must excuse me, I am a
working man... Mine is fine work. I must have damages, for I shan’t be able to use this finger
for a week, maybe... It’s not even the law, your honour inspector, that one should put up with
it from a beast... If everyone is going to be bitten, life on earth won’t be worth living...”
“Hm!.. Good...” says Otsjoemelov Jansen sternly, coughing and raising his eyebrows.
“Good... Whose dog is it? I won’t let this pass! I’ll teach you to let their dogs run all over the
give him a lesson!.. Jeldyrin De Vries,” says the superintendent, addressing the policeman,
“find out whose dog this is and draw up a report! And the dog must be destroyed. Without
delay! It’s sure to be mad... Whose dog is it? I ask!”
“I fancy it’s General Zjigalov’s chief inspector Jacobs’,” says someone in the crowd.
“General Zjigalov’s Chief inspector Jacobs’? Hm!.. Help me off with my coat,
Jeldyrin De Vries... it’s frightfully hot! It must be a sign of rain... There’s one thing I can’t
make out, how it came to bite you?” Oetsjoemelov Jansen asks Chrjoekin Tiggelaar.
“Surely he couldn’t reach your finger. He’s a little dog, and you are a great hulking fellow!
You must have scratched your finger with a nail, and then the idea struck you to get damages
for it. We all know... your sort! I know you devils!”
59
“He put a cigarette in its face, your honour inspector, for a joke,” says someone in
the crowd. “But but it had the sense to snap at him... He is a nonsensical fellow, your honour
inspector!”
“That’s a lie, Squinteye!” Tiggelaar calls out. “You didn’t see, so why tell lies about
it? His honour The inspector is a wise gentleman, and will see who is telling lies and who is
telling the truth, as in God’s sight... And if I am lying let the court decide. It’s written in the
law... We are all equal nowadays. My own brother is in the gendarmes... if you want to
know...”
“Shut your mouth!” Jansen orders.
“No, that’s not the General’s chief inspector’s dog,” says the policeman De Vries,
with profound conviction, “the General chief inspector hasn’t got ones like that. His are
mostly setters.”
“Do you know that for a fact?” Jansen asks.
“Yes, your honour boss.”
“I know it, too.,” says Jansen. “The chief inspector The General has valuable dogs,
thoroughbred, and this is goodness knows what! No coat, no shape... a low creature... And to
keep a dog like that?!... Where’s the sense of it. If a dog like that were to turn up in
Petersburg or Moscow, do you know what would happen? They would not worry about the
law, but at once – away with it! You’ve been injured, Chrjoekin Tiggelaar, and we can’t let
the matter drop... We must give them a lesson! It is high time...”
“Yet maybe it is the General’s chief inspector’s...” says the policeman De Vries,
thinking aloud. “It’s not written on its face... I saw one like it the other day in his yard.”
“It’s the General’s chief inspector’s, that’s certain! “ says a voice in the crowd.
“Hm!.. help me on with my overcoat, Jeldyrin De Vries, my lad...” Jansen requests
De Vries. “The wind’s getting up... I am shivering... Take the dog to the General’s chief
60
inspector, and inquire there. Say I found it and sent you... And tell them not to let it out into
the street... It may be a valuable dog, and if every swine goes sticking a cigar to its nose, it
will soon be ruined. A dog is a delicate animal... And you Tiggelaar, put your hand down,
you blockhead! It’s no use your displaying your fool of a finger! It’s your own fault!....”
“Here comes the General’s cook chief inspector’s gardener, ask him...” De Vries
remarks. “Hi, Prochor Nederhof! Come here, my dear man! Do you see that dog?... Is it
yours?”
“Are you kidding? We have never had one like that!” says Nederhof.
“There’s no need to waste time asking,” says Otsjoemelov Jansen. “It’s a stray dog!
There’s no need to waste time talking about it... If I say it’s a stray dog, a stray dog it is... It
must be destroyed, that’s all about it.”
“It is not our dog,” Prochor Nederhof goes on. “It belongs to the General’s chief
inspector’s brother, who arrived the other day. Our master chief inspector does not care for
hounds. But his honour is fond of them...”
“You don’t say his brother has arrived? Vladimir Ivanytsj Cas Jacobs?” inquires
Otsjoemelov Jansen, and his whole face beams with an ecstatic smile. “Well, I never! And I
didn’t know! Has he come on a visit?
“Yes...” Nederhof starts.
“Well, I never... He was missing his brother...” Jansen continues. “And there I
didn’t know! So this is his honour’s dog? Delighted to hear it... Take it... Good dog... A
lively creature. . . . Snapped at this fellow’s finger! Ha-ha-ha... Come on, why are you
shivering? Grrr... Grrr... The rogue’s angry... a nice little pup.”
Prochor Nederhof calls the dog and walks away from the timber-yard with it. The
crowd laughs at Chrjoekin Tiggelaar.
61
“I’ll get you!” Otsjoemelov Jansen threatens him, and wrapping himself in his
greatcoat, goes on his way across the square.
62
True Colours, non-fictional text (English translation provided here, Dutch original used)
There are but few animals with such an extensive collection of anatomical curiosities as the
chameleon. It has a tongue that is longer than its body; with it, it can attack an insect in a
fraction of a second. It also has telescopic eyes in round sockets that can move independently.
It can use its toes as a pair of tweezers, has lumps on its forehead and snout, strange bulges
on its nose, and it has a skinfold in its neck that resembles the lace collar of a sixteenth
century nobleman.
Its most typical trait amazed people as far back as the days of Aristotle: its ability to
change colour. It is a widespread myth that the chameleon takes on the colour of everything it
touches. Sometimes such a change of colour does make him stand out less in his
environment, but in reality it is a physiological reaction that serves as a means of
communication. By means of colour, the lizard expresses everything it feels, like attraction,
rivalry, or threat from its environment.
At least, that is the current view. “Even though chameleons have interested people for
years, they are still surrounded by a lot of mystery”, says Christopher Anderson, biologist and
chameleon expert of Brown University in Rhode Island. “We still don’t know how their
mechanisms work” – from the rapid tongue movements to the changing skin colour.
In recent years, important discoveries have been made through observations of chameleons in
captivity. Their future in the wild is however highly uncertain.
According to the most recent Red List of the International Union for Conservation of
Nature (IUCN), that was published November last year, at least half of all chameleon species
are considered to be ‘endangered’ or ‘near endangered’. Anderson is part of the chameleon
work group of the IUCN, just as biologist Krystal Tolley. She does field work in the south of
Africa, where she has discovered new chameleon species and is mapping their disappearing
habitat.
63
Approximately 40 percent of the over two hundred known chameleon species lives on
the island of Madagascar, most others in continental Africa. Over 20 percent of the known
species has only been discovered in the past fifteen years.
Anderson closely studies the feeding behaviour of the chameleon. Using a camera that
captures three thousand frames per second, he has shot 0.56 seconds of footage of a
chameleon eating a cricket. He has stretched out that footage to create an instruction video of
28 seconds, which shows exactly how the animal uses its tongue.
In its throat sac the chameleon has a tongue-bone surrounded by elastic connective tissue in a
tubular acceleration-muscle. As soon as the chameleon sees an insect, it sticks out its tongue
slightly, which causes the muscle to contract and the tongue to shoot out as if it is launched
by a catapult. The tip of the tongue functions as a suction cup that grabs onto the prey, after
which the tongue slides back inside.
But we are far from knowing everything about the mechanism of the tongue, says
Anderson. His research suggests that some chameleons can extend their tongue even faster
and farther than was previously thought.
Also with regards to the changing skin colour there are new insights – especially since
the publication earlier this year of the research of Michel Milinkovitch, evolutionary
geneticist and biophysicist at the Université de Genève. For a long time, the dominant theory
was that the chameleon changes its colour by effusing pigment cells in the skin through
paper-thin cell extensions. Milinkovitch does not buy this theory, because many green
chameleons don’t even have green pigments in their skin.
That is why he decided to “bring physics and biology together” in his team, he says.
Under a layer of skin with pigment cells they discovered another layer with cells that contain
nanocrystals, ordered in a triangular grid.
64
By exposing skin samples of chameleons to pressure and chemicals, the scientists
discovered that the pattern of crystals can be adjusted by increasing or decreasing the distance
between the crystals. And that has an effect on the light that the crystals reflect. With an
increase of the distance between the crystals, the reflection of light changes from blue to
green, yellow, and orange and finally to red– the kaleidoscopic colour show that can be
clearly seen on some panther chameleons when a calm disposition changes to aggression or
arousal.
According to a study from 2014, chameleons have also developed the ability to show
submission through the changing of their skin colour, because due to their “slow tempo they
are hardly capable of getting away quickly from dominant conspecifics”.
Although all chameleons change colour, some species do not do so in a way that is
spectacular enough to scare away their rivals. Fortunately the animals have another
intimidation technique: they can make themselves appear bigger. They make their body
longer and higher by expanding their ribcage. They can also make themselves more massive
by rolling up their tail and using their tongue to puff out their throat.
To escape the birds and snakes that hunt them, the chameleon has developed new
ways to make itself invisible. Most species live in trees; if they make themselves narrow they
can hide behind a branch. If species that live on the ground see a predator, they disguise
themselves as a leaf by contorting their body in such a way that they resemble crumpled
leaves on the forest floor.
But there are some dangers the chameleon cannot escape, for example when the forest
it lives in is burned down to make place for agriculture. According to the IUCN list, 9 species
are ‘critically endangered’, 37 ‘endangered’, 20 ‘vulnerable’ and 35 ‘near endangered’.
Since 2006, Tolley has discovered eleven new chameleon species in South Africa,
Mozambique, Tanzania and the Democratic Republic of the Congo. “If DNA research
65
confirms that it is indeed a new species, you feel like you are not just writing a scientific
article that no-one reads anyway”, Tolley says. “Then you have achieved something that
lasts.”
But then she immediately says: “At the same time you think: wow, this is amazing, it
strikes terror in me. Because I constantly imagine how those little chameleons clamp onto
their branches while the forest around them is destroyed.”
Hoarsely: “And then I wish we’d never found them. Because if this does not end, they
will soon be extinct.”
i
Unfortunately, experimenter error entailed that one first order cognitive trial too many was included. The
original version of the task thus has 12 first order cognition trials instead of 13.
ii
It should be noted that the outcomes for these additional trials were not inspected at any point in the
preparation of this manuscript. In keeping with the Bayesian analysis approach, only those analyses were
conducted for which hypotheses had been specified prior to data analysis (deviations from this approach are
clearly marked as such in the Data Analysis section). The decision to include only the core Yoni trials in this
study was thus based only on the fact that these were the trials used in previous related research in which the
Yoni task was applied and on no other considerations.

iii
Note that separate analyses were run for the two different ToM measures, but the predictions are the same for
both measures.
iv
None of the participants indicated that they were already familiar with the texts that they had read, so none
were excluded on this basis.

v
For the unconstrained hypothesis a vague prior was used and the priors for the other hypotheses are derived
from this vague prior via truncation. Discussion of this issue is very technical in nature and beyond the scope of
this paper. However, the reader is referred to Mulder, Hoijtink & Klugkist (2009) and Mulder et al. (2009, 2012)
for further discussion.

vi
It should be noted that the level of uncertainty that is present in the data is reflected in these PMP values. That
is, more uncertainty in the data leads to lower PMPs for the pre-specified hypotheses and makes it less likely
66
that one hypothesis will clearly come out as the ‘winner’. Giving an overview of the confidence intervals that
are associated with the data thus does not provide any additional insight above and beyond the PMPs.
vii
The same analyses were also run controlling for ART Overall scores, but these differ only very minimally
from the analyses provided in Table 5, given that the differences lie in the ART Literary scores.
viii
Instead, they have focused on including narrative fiction authors from various different genres, as well as
distinguishing fiction from non-fiction authors.
67
View publication stats

LiteratureToM Resub Copyedited

Uploaded by

Copyright:

Available Formats

LiteratureToM Resub Copyedited

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

LiteratureToM Resub Copyedited

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Effects of exposure to literary narrative ﬁction: From book smart to street

Article in Scientific Study of Literature · November 2017

Hannah De Mulder Frank Hakemulder

SEE PROFILE SEE PROFILE

Rianne van den Berghe Fayette Klaassen

L2TOR - Second language learning with social robots View project

“Empirical Study of Literature Training Network” (ELIT) View project

The user has requested enhancement of the downloaded file.

Please cite as:

Running head: EFFECTS OF EXPOSURE TO LITERARY NARRATIVE FICTION

Hannah De Mulder is now at LUCL, Leiden University, The Netherlands

for his help with programming the tasks.

associated with more advanced ToM.

recognition test, Bayesian analyses

Narrative Fiction and Understanding Others’ Mental States

which it was possible to tie a rope.” (p. 1; Eugenides, 1993)

Zunshine, 2006). Social-cognitive processing is thus necessary in order to be able to

experiencing them ourselves.

& Mazza, 2016).

Romance, Science-Fiction/Fantasy and Suspense/Thriller) and ToM task performance.

exposure to Romance fiction being responsible for the ToM enhancement.

generally considered to be an instantiation of literary fiction). Furthermore, it is also not

narrative fiction in enhancing our understanding of others. Potentially, literary narrative

subjective experiences in order for full understanding to be achieved. Furthermore, literary

& Hakemulder, 2015). In contrast, non-literary narrative fiction is considered to provide a

That this interpretation of previous findings may be correct is supported by a study

being provided with the content of others’ mental states explicitly.

demonstrates an awareness of how their emotional response to something transformed over

skills (cumulative effect).

Literary Narrative Fiction and ToM: The Relevance of ‘Gaps’

In Iser’s conceptualization of information gaps it is assumed that these are characteristic of

differences between the groups could be determined.

(depending on the extent of lifetime literary narrative fiction exposure).

This design thus enables us to contribute to current knowledge on the relationship

specific to, literary narrative fiction.

Competing Hypotheses Regarding Effects of Narrative Fiction Exposure on ToM

assumptions behind them are detailed here.

To investigate direct effects of exposure to (literary) narrative fiction on ToM ability

evaluated hypothesis is based on a more extreme interpretation of Kidd and Castano’s

To investigate cumulative effects of exposure to (literary) narrative fiction on ToM

characteristic of the story A chameleon is that it remains implicit which character is

extensive manipulation version would be provided with fewer opportunities to engage in

Examples of the original text and the two manipulated versions

Text type Text example

Original “Shut your mouth!”

“Do you know that for a fact?”

Limited manipulation “Shut your mouth!” Otsjoemelov orders.

“Do you know that for a fact?” Otsjoemelov asks.

Extensive manipulation “Shut your mouth!” Jansen orders.

“No, that’s not the chief commissioner’s dog” says policeman De

Vries with profound conviction, “The chief commissioner hasn’t got

ones like that. His are mostly setters.”

“Do you know that for a fact?” Jansen asks.

Note. Differences as compared to original text are rendered in bold face.

regarding reading frequency/duration. Attempts to assess general print exposure have

Although we acknowledge that we have no direct validation of the relationship

there were also made-up names in the list.

is being signalled. Although some of the response options refer to an understanding of