WHAT LITERATURE DOES TO OUR EMOTIONS, AND HOW DO WE KNOW? EMPIRICAL STUDIES WILL TELL

The general feeling of malaise, if not crisis, in Literary Studies forces us to urgently look for solutions that will bring the discipline forward. This article is a call for a concentration on fundamental issues in the study of literature, and at the same time for a more rigorous and accountable methodology in studying both the content and the form of literary texts as well as readers’ reactions to them. Some illustrations of work in the area of Empirical Study of Literature are provided, showing how fiction is a powerful regulator of human emotion, especially by formal features of the text. Case 1 reports a study which looks at the influence of narrative perspective (internal focalization in the first place) on judgements of readers. Case 2 delves into the textual ingredients by which readers’ absorption in a narrative world is enhanced. These ingredients are foremost of a kind that goes under the name of “foregrounding” devices in literary studies. The conclusion from the research is that texts that are rich in foregrounding are better able to elicit a more complex response, i.e., a more powerful impact, from readers. In its turn Case 3 looks at how readers react to literary pieces dealing with deep human suffering. The findings indicate that literature is able to evoke strong feelings of empathy through its formal make-up. The results also support the argument that one’s exposure to literature is the main variable to have an impact on prosocial behaviour, irrespective of personality, gender, age or social situation. Thus we claim that literary texts exert a powerful influence on readers’ value sharing, absorption and empathy, and the impact can only be studied empirically. The article shows a way out of the current crisis, not by just opening up a new fashion, in which literary texts are “interpreted” in yet another way, mostly academically, but by taking literary texts seriously in their workings on the minds and hearts of readers — which is ultimately what texts are written for.


An on-going crisis in literary studies
It is now almost fifteen years ago that a conference took place at the University of Chicago, dealing with the current state of literary studies. (See Mitchell, 2004 for a report in the famous journal Critical Inquiry). Several celebrities from the field dealt with the future of the discipline, among others: Stanley Fish, Henry Louis Gates, Jr., Hillis Miller, Fredric Jameson and Homi Bhabha, all star-studded names in the field of what has become called : "literary theory". The general mood was one of malaise, if not severe depression, expressed most lucidly by Jerome McGann in his opening statement: "A widespread malaise has been notable in our discipline for more than a decade" (McGann, 2004, p. 409). If that statement has its time line correct, then we are currently facing a deep crisis in literary study for almost a quarter of a century now. Or longer even. As Menand (2010, p. 62) points out: "In the 1980s, people began wondering what the return on investment was in the humanities". The question that looms over this predicament is whether the crisis has been overcome. That does not seem to be the case, though the celebrities have stopped worrying in public. As Gottschall recently observed: It seems that literary scholars are to be the laughingstocks of the academic world (see Delbanco, 1999, p. 32;Pattai & Corrall, 2005, p. 18;Oakley, 1997, p. 67). We are savagely parodied in academic novels, humiliated by hoaxers, and held up to ridicule by satirical journalists (…) This is all revenge for our perceived pretentiousness, for the impenetrability of our verbiage, for our unearned moral vanity, and for our apparent contempt for reality (2008, p. 1).
The causes for this sad state of affairs may be manifold. Gottschall (2008) referred to a lack of methodological training. As Menand has observed, there is also "no clear agreement on a definition of what humanists do" (2010, p. 63). And, despite their stable popularity in Eastern Europe, post-Soviet countries in particular, the numbers of majors in English and the Humanities are drastically falling nationwide in the US: see the recent data from the Integrated Post Secondary Education Data System: https://www.insidehighered.com/ news/2015/01/26/where-have-all-english-majorsgone (consulted 16.02.2019). This is all the more sad, because insight into the value of the Humanities is of vital importance for understanding our own culture; for a magisterial overview, see Bod (2016), who also argues in favour of close cooperation of the Humanities and the sciences.
This lack of focus is compounded by a near-total lack of non-partisanship. According to Gross & Simmons (2007), in their in-depth study of the world views of Professors at American institutions of higher education, report that 71 % of their respondents found it OK that professors let their research be determined by their political or religious beliefs. Quite frightening, furthermore, is that only 5 % of professors in the liberal arts describe themselves as "ardent advocates of neutrality". Take heed: barely 5 % of American university professors in the Humanities subscribe to the idea that knowledge (which is their profession) should be neutral. This is the surest and fastest remedy toward self-abolition.
And we are currently witnessing the results: in a recent contribution to the New York Review of Books, the Shakespeare specialist James Shapiro of Columbia University concludes: With funding for higher education slashed, literature departments downsized, full-time faculty replaced by adjuncts, and illustrious universities like my own choosing to hire only at the entry level to replace those of us who will be retiring, the prospects facing the next generation of academics are dismal. Depressingly, there is only a single position advertised this year in all of North America for a senior Shakespeare scholar (April 19, 2018).
Just imagine: in the vast continent of North America, with immense financial means and a strong will to outperform other nations, there is no prospect whatsoever for young scholars in the field of literary studies. In Europe, we have not woken yet from the dream that we can go on as usual. But we will soon enough be woken up. This alerts us to urgently request our colleagues to wake up: we are utterly naïve driving a fast car blindly in the direction of an abyss. But colleagues in literary studies urge us to go on as usual, as if nothing has happened and nothing IS happening. Soon they will notice that they will have no chairs, no positions, nor Ph.D. students, because there WILL be no future.
What is to be done? What would be more natural than to look for a new paradigm as a solution? But, alas, that will not work. Because this is exactly what has been tried. Since the 1980s various proposals in such a direction have been made, be they Postmodernism, Deconstruction, New Historicism, Feminism, Queer Theory, and so forth. What has transpired over the past decades was that this wish for a new paradigm was but a chimera. What we have seen instead is a series of substitutions: one fashion has been replaced by another one, ever more promising than the preceding one, without ever fulfilling any promises before it was ousted by the next fashion. In fact, this has aggravated the situation, first of all by augmenting the confusion, but also in that it created the illusion of progress: every new "paradigm" seemed to supersede the previous one, while in reality it replaced it, without making any progress. The process is indeed, in the very sense of the word, the same as in fashion: colours and shapes come and go, without any rationale.

What is to be done?
We have dwelt somewhat longer on the regrettable state of affairs, because it is dramatic, not just in terms of the Humanities' position in society, but more so even in the light of the potential that these disciplines could bring to a deeper understanding of what it means to be human. The above may raise the question of whether we may not differ in viewpoint, in theoretical background, or in our various approaches. The simple answer is not that we may, but that we must, because critical confrontation is at the heart of research. Unfortunately, doing research together with students is not practised frequently in the Humanities. A good deal of teaching consists in reading some theoretical material -and then "applying" this method to the interpretation of texts. But that is not research. It is (most often naïve) interpretation.
In this article we are going to suggest an alternative to our current situation. We sincerely believe that it offers a better perspective for the future of our field for a number of reasons. Our belief is grounded in evidence: that students who go through this new approach find good (that is: interesting and well-paid) employment on the job market. We can claim this, looking at past practises over a period of some three decades now. And we are sure that going on like we do at present, doing more of the same will destroy the future. It will bring literary studies to a grievous end.
So what is our proposal and what are its advantages? Our proposal is to, at least in part, get involved in empirical research into literature, not necessarily in our own research, but in class, i.e., with our students. There is a massive gap in our understanding of what literature does to readers: why they read in the first place, what emotions are evoked by which kinds of literary content / structure, how literature influences people's lives, etc. All of these questions have been and are frequently asked. In order to find answers to them, one will have to look at real readers, either in the form of documents (testimonies, diaries or correspondence) or by asking or observing readers "in the flesh". Of course, an empirical approach should not be the only one in literary studies, but those questions can only be answered through empirical research, not through hypothetical speculation. Nor through hermeneutic or historical methods. Both of these methods have their place and their validity in the study of literature, but functional questions about motivations to read and effects of reading can only be answered through the use of empirical research methodology.
This proposal does not imply that current elements of the literary curriculum have to be abolished. Quite the contrary: as this article will show, there is an added value if such traditional elements can be coupled to reader research. We will further give, as an example, the way in which the study of narrative structures can be fruitfully linked to investigations of real readers, thereby actually enriching narratology. Reader research that is not coupled to insights from literary studies often remains sterile. Our proposal is aimed at mutually profitable interactions between current issues in literary studies and investigations of reading processes that are relevant to it.
There are, in our experience of using these methods, considerable benefits. Studying literature now becomes problem oriented, demanding students' active participation in conceptualizing and carrying out research themselves, which always gives an enormous boost to their motivation. It is also a highly active form of learning, both in small teams with fellow students and with their lecturers. It leads to concrete results, boosting their satisfaction with their work. The fact that this approach is highly interdisciplinary, usually adds to the flavour in the eyes of the students.
In this way, Chesnokova, Zyngier & van Peer (2017) actually did research on a ten-year international empirical project with students from three different cultures: Brazil, Germany, and Ukraine. The results are revealing: in general these students left university with the outcome of research they had conceived and carried out themselves, of which they were proud, and which they could show to future employers, heightening their positions on the job market. They indeed all got excellent jobs, but not just in literature or at the university, but also in government, industry, international organisations, administration, or business. The research project showed that these students had developed a high level of skills in problem solving, teamwork, and sophisticated methods of data analysis. But perhaps even more important was the fact that they had developed a much keener awareness of the social significance of the Humanities. They did not study literary texts for an exam, but had internalized and integrated into their personality the power and purpose of literature.
At this point some might object that students are simply not able yet to carry out research though that is not the experience of most teachers in empirical studies. And van Peer, Hakemulder & Zyngier (2012) is a hands-on introduction to guide them through the process. This process of binding students into the research process is standard in most natural sciences. If our colleagues in these fields can accomplish this, we would cast doubt on our pedagogical faculties.
If a consensus on rule-based methodology and evidence based research in literary studies could be established, quite some progress might be achieved. With one proviso: that the discussions are geared toward real problems in the field. If they are not, they may have all the appearance of academic quality, but will remain just this indeed: "academic"; see van Peer (2008) for an incisive criticism of the Humanities and its detachment from human concerns. The philosopher of science Mario Bunge once made fun of such academism when pointing out that "a comment on Cicero's discussion of Clitomachus' account of Carneades' views is likely to be regarded as the summit of serious scholarship" (Bunge, 2001, p. 208). Or, as James Rule commented: no research paradigm "[w]ill long claim attention if it fails to address a core of historically enduring questions" (1997, p. 45). Not questions like: "Was Ukraine at the end of the 16th century part of the 'periphery' of Europe, or of its 'semi-periphery'?" These are just funny examples of the kind of academism that sometimes prevails in our field too. In what follows we wish to concentrate on one issue that we believe to belong to the core of such real problems in literature. It is the role emotions play in reading or listening to literary texts.

Emotions in Literary Research
The efforts to study emotions in literary studies gyrate around two poles. The one is the description and evocation of emotions in literary texts, for instance, what emotions characters in a story experience. The other is the emotional reactions of readers and spectators to these emotional evocations by texts. It is not that these are totally new topics, for both aspects draw attention in literary studies. If literary studies neglect the emotional impact of reading or listening to literature, it forfeits one of its central concerns. There may be many reasons why people read literature: to gain knowledge, to become acquainted with a particular author, to escape daily concerns, and so forth. Yet some kind of emotional gratification is often involved when people read poetry, novels or plays.
But what is it in literary texts that moves us? At first sight, one might think of words that are emotionally coloured that evoke emotions in the reader, and that may well exert some influence over readers' reactions. But ever since Miall & Kuiken (1994) published their findings, we know that this is only a small part of the truth. It emerged that not so much emotional words, but "non-emotional" linguistic structures provoked strong feelings in the reader. Such linguistic features are described by the theory of foregrounding. It found its most eloquent expression in Victor Shklovsky's (1917Shklovsky's ( /1965  And art exists that one may recover the sensation of life, it exists to make one feel things, to make the stone stony. The purpose of art is to impart the sensation of things as they are perceived and not as they are known. The technique of art is to make objects 'unfamiliar', to make forms difficult, to increase the difficulty and length of perception because the process of perception is an aesthetic end in itself and must be prolonged. Art is a way of experiencing the artfulness of an object: the object is not important (idem, p. 12).
For further elucidation of the concept of foregrounding, see Leech (1969), Leech & Short (2007), Short (1996), Simpson (2014), van Peer (1986), , van Peer,  and Wales (2001). We are dealing with a theory in literary studies that has been around for a hundred years now, and which is one of the very few theories that have been subjected to rigorous empirical testing; see, for instance, Hakemulder (2008) (2007), van Peer & Nousi (2006), the 2007 issue of the journal Language and Literature and more recently Meyer & Berlina (2018).
In general, the underlying model for testing a theory can be rendered graphically as follows (see Fig. 1 below): The original claims by the foregrounding theory, as formulated by Jakobson (1960), Mukařovský (1965), and Shklovsky (1917Shklovsky ( /19651929/1990, are cast in the form of predictions, which are then matched to independently collected data. These data are subjected to a statistical analysis -first of all, to see whether they are in the predicted direction and whether they are convincing enough to be accepted and generalized. This is expressed in a p-value, which stands for error probability, which should, naturally, be as low as possible, in any case lower than .05 (which equals roughly to 5 %). The procedure and rationale for doing so is explained in detail in van Peer et al. (2012). Results in the predicted direction that have error probabilities lower than 5 % (p-values lower than .05) are then taken as support for the theory in question. At the same time, recently this picture has been criticized, along with a recommendation to attach more meaning to estimations of effect size, confidence intervals, and meta-analysis; see Cumming (2013).
What follows are three examples of how the workings of emotions in the act of reading can be investigated.

Case Study 1: Emotions and Narrative Perspective
Everyone is familiar with the notion that stories can be told from different points of view. But why are there such differences? Presumably narrative perspective must serve some function. It is not just the difference between a 1 st or 3 rd person narration (to mention the two most common ones), but also the kind of information the narrator is prepared to provide (or not). Internal focalization is such a technique: it gives readers information about characters' (sometimes also narrators') thoughts and emotions. It is assumed in narratology that this will have an effect on the reader. In this respect Leech & Short propose the following "theory": the very exposure (…) to a character's point of view -his thoughts, emotions, experiencetends to establish an identification with that character, and an alignment with his value picture (2007: 221).
The claim is clear enough. But is it true? Is it indeed the case that internal focalization brings about the sharing of values with a character by the reader? Intuitively this seems plausible. But as we know from experience, intuitions can be deceptive. The sun does not rise in the morning, in spite of our very strong intuitions saying so. If we wish to investigate this little "theory" in a more rigorous way, as proposed in the previous sections, we will have to follow the model outlined above. (The full report of this research is in van Peer, 2001).
The first step is to derive predictions from the statement by Leech and Short. One could argue that an alignment with the value concerns of a character translates into some kind of sympathy for, or empathy with, that character. What would this mean in practice?
To investigate such a claim, a story was used in which a conflict between spouses erupts in the course of events. A husband is watching sports events on television while his wife is washing up in the kitchen. The story is told by a neutral narrator, who merely observes and records the "facts". Thus the story starts as follows: It is evening. Ted Jones, factory worker, is sitting in front of the television set, watching the sports events, while his wife, Liz, has just finished washing up in the kitchen.
"Turn down the volume," she says, "I want to talk to you." Ted does as is asked. In order to demonstrate the effects of internal focalization, two different versions of the story were created, one in which thoughts and emotions by the female character were inserted, and the other one containing thoughts and emotions by the male character -in exactly the same text locations. Thus the female version ran as follows: "Turn down the volume," she says, "I want to talk to you." "Can't he ever think of that himself?" she thinks.
Ted does as is asked.
While the male version ran: "Turn down the volume," she says, "I want to talk to you." "Can't she ever leave me alone?" he thinks.
Ted does as is asked.
As can be seen, the manipulation was minimal, and displays almost identical syntax. (The new versions were somewhat longer than the original, of course, due to the insertion of the internal focalizations.) To check the effect of the perspective change, the three versions ("neutral", "male" and "female") were distributed randomly among readers who were requested to indicate (on a scale) how "just" and also how "considerate" they found each character. The results for the first question can be seen in Graph 1 below: Graph 1. Evaluation of characters across the three versions: the "just" variable "Liz" and "Ted" are the names of the respective female and male characters. Compare the perceived sense of being "just" by readers allocated to the two characters. As can be seen from the graph, Ted scores much higher by readers who had read the version in which thoughts and feelings of the male character were inserted, while the reverse is the case for the "female" version: in line with the predictions, Liz scores highest in perceived justness by those readers.
So far one could surmise a corroboration of the theory as propounded by Leech and Short. But look at the bars in the middle: they represent the scores by readers who read the "neutral" version, i.e., the one without any internal focalization. Intuitively one would expect those readers' perception to be somewhere in between the "male" and "female" versions, but that is not the case. In fact, readers who read the neutral version judged the female character considerably more just than readers who had read the "female" version, i.e., the version in which internal focalizations of the female character had been inserted. This is not right, according to the theory. But it is what readers felt.
The problem becomes even more apparent when we look at the way in which participants judged the protagonists' attitude of being considerate in the conflict. For the results see Graph 2 below: Graph 2. Evaluation of characters across the three versions: the "considerate" variable As shown in the graph, for Ted, the male character, the inserted focalizations exerted an effect, albeit a modest one. But there was no difference in the reactions to the female character's being seen as considerate depending on the focalizations: even the version with the negative feelings of the male character scored just as high as the one with female emotions. And again the neutral version scored higher than the two manipulated ones. Notice also that the female protagonist is judged quite lower than the male one in all three versions.
The conclusion from the study must be that there is some truth to the proposal by Leech & Short (2007) which very much reflects the outlook of narratology about the effects of internal focalization. At the same time, however, the results show that there are serious difficulties with this theoretical outlook. The fact that focalizations had no effect whatsoever in the judgement of the female character's considerateness, directly goes against the predictions of the theory. Readers apparently experience emotions that are not (or wrongly) predicted by the theory about internal focalization.
Here we encounter a major advantage of the empirical method proposed here. It allows us to detectin a straightforward and rather simple way -where our theories are deficient. This does not automatically force us to abandon the theory, but it does cast some serious doubt over its generality. Obviously, other factors than focalization are also at play in readers' emotional reactions. Without taking care of such factors, the theory will lose a considerable portion of its explanatory power.

Case Study 2: Absorption and Foregrounding
We have already mentioned Miall and Kuiken's path-breaking insight in readers' emotional reactions to foregrounding stretches of text that do not contain any emotional load in themselves. Maybe this should be taken as less surprising than it appears at first sight. After all, we may get chills or gooseflesh on hearing a piece of music. But the notes in themselves have no meaning. And yet they may evoke powerful feelings in listeners.
In an in-depth study of absorption Kuijpers (2014) has further expanded this line of thought. Absorption is an emotional state widely acknowledged by readers, in which everyday concerns recede totally into the background of consciousness. Most people know the experience of being absorbed in a story. It means living "in" the world of the story during reading, engaging with the characters and events, of being unaware of one's surroundings and daily preoccupations. The phenomenon has been the subject of various investigations. There is now a considerable body of research probing this type of emotional involvement with narratives, albeit under different terms: transportation, identification, narrative presence, and, indeed, absorption. Major studies in this area are Bussele & Bilandzic (2009), Cohen (2001, Gerrig (1993), Green & Brock (2000), Nell (1988), Slater & Rouner (2002) and Tal-Or & Cohen (2010). Most recently a volume (Hakemulder, Kuijpers, Tan, Balint, & Doicaru, 2017) containing a collection of fourteen research reports has been added to the list, so there is now a considerable body of research studying these emotional kinds of involvement with literary texts from an empirical vantage point.
Many of these studies probe what may be loosely called the "content" of stories. Kuijpers (2014), however, delved into the concrete language by which absorption was called forth and thus investigated the purely textual ingredients by which absorption in a narrative world comes into being. She looked at both popular narrative devices and more "literary" ones, such as foregrounding. Here unmistakable individual differences could be observed, and Kuijpers emphasizes that these should not be ignored in research. But such differences do not lend credit to postmodern or constructivist theories claiming that readers construct their own individual texts. Quite the contrary: by meticulously scrutinizing textual structure, the results prohibited any such fashionable postmodern assertions. It is the story that determines whether and in what way experiential states contributed to participants' emotional responses. Moreover, on the basis of previous encounters with narratives, Kuijpers found that [p]rint exposure was found to have a negative significant effect on the relationship between text and story world absorption when using popular textual devices. (…) Readers who have had high print exposure experience more story world absorption in texts that make use of deviation (idem, p. 154).
This again underlines the preponderance of formal structures of the text: foregrounding was a predictor of both enjoyment and impact (…) impact is seen as a more complex evaluative response than enjoyment. (...) Foregrounding being an experiential state that includes reflection, is perhaps better able to elicit a more complex response such as impact" (idem, p. 153).

Case Study 3: Literature and Suffering
The emotions depicted in literary texts are generally of a positive kind: "Clearly, literature of all kinds gives more weight to positive than to negative emotions, and has done so consistently for as long as words have been written" (Lindauer, 2009, p. 84). On the other hand, from its very beginnings, at the dawn of human civilization, literature has also probed the depths of human suffering. In the oldest extant literary work of humanity, The Epic of Gilgamesh, the hero complains bitterly of his friend's death. And it is not difficult to compose a list of major works of literature that centre around human suffering: Steinbeck's The Grapes of Wrath, Tolstoy's The Death of Ivan Ilyich, or Hugo's Les Misérables, to name only a few. The same holds true for national texts examples, such as Ukrainian Bahrianyi's Garden of Gethsemane or Polish Dąbrowska's Nights and Days. Though positive emotions may generally emanate from literature, the number of works dealing with deep human suffering matches the more pleasant feelings if not in quantity then certainly in acuteness and profundity. Why this is so, is a paradox: why do people wallow in descriptions of misery, hardships and distress? Presumably, the emotions induced by such confrontations with suffering must be unpleasant. So why? So far, the paradox is unresolved, although some recent empirical work, notably by Menninghaus, Wagner, Hanich, Wassiliwizky, Jacobsen, & Koelsch (2017) and van Peer, Chesnokova, & Springer (2017), has begun to unravel the mystery.
Recently Koopman (2011; has added to this in her study of suffering in literature. What she found was that readers say they are first and foremost looking for "meta-emotions": the need for feeling, liking to feel emotions while reading. But on top of that, and specifically with regard to suffering, they reported looking especially for "insight" and "personal growth". Surprisingly, no gender effect was observed though it could be, of course, that male readers may differ from the general male population. What is perhaps most striking in Koopman's findings is the fact that, as with Miall & Kuiken (1994a) and Kuijpers (2014), the formal make-up of the language plays a dominant role. Let us now concentrate on this aspect of her research for a moment. Koopman had volunteers read a story about deep emotional suffering (the loss of a child). The story contains heavy foregrounding, as defined before (see Section 4 of this article). Koopman then tampered with such foregrounding, in that she removed it and replaced it by conventional, everyday language. She produced, next to the original text, a version with moderate and one with no foregrounding at all. These versions were read by three groups of different, but in all respects highly comparable, groups of readers. After having read the story, they were requested to answer questions dealing with the degree of empathy they felt for the protagonists. This means, in other words, that all participants in the experiment read about the same characters, the same events, the same feelings, and so forth. The only difference lay in the language that was used: from everyday language to highly foregrounded one. Intuitively one would surmise that colloquial language, being closest to readers' daily experiences, would produce the strongest feelings of empathy. But the reverse was the case: it was the version containing the form most remote from participants' common language that caused the strongest participatory emotions: The foregrounding study, however, which compared reactions to three versions of one literary text differing in foregrounding did [show the differences on an attitudinal measure of empathy]: those readers who had read the most original text (containing most foregrounding) had the highest empathic understanding afterwards. This effect was upheld when controlling for personal factors like trait empathy (Koopman, 2016, p. 239-240).
And this effect was noticeable after reading only 1,500 words! The final sentence in this quotation is important. One could presume that it is precisely readers who of themselves are already strongly inclined toward feelings of solidarity (with someone suffering) who would experience the strongest feelings of empathy. And would therefore be less influenced by the concrete form of language. But this was not the case: "trait empathy" refers to the personality characteristic that distinguishes people's empathic inclinations. This personality trait did not make a difference: it was the language that did! Thus literature is able to call forth strong feelings of empathy through its formal make-up. It also turned out that one's exposure to literature was the only variable to have an impact on prosocial behaviour, again irrespective of gender, age or social situation.

Conclusions and Discussion
The previous sections have provided illustrations of the way literature evokes emotions, and how readers deal with them. These examples could be augmented with many more, also from the empirical methodology that has been advanced here. There is an international society concentrating of this kind of research, IGEL 1 , the International Society for the Empirical Study of Literature (see the website at https://sites.google.com/igelassoc.org/igel2018/home), with also a journal, Scientific Study of Literature (see https://benjamins.com/#catalog/journals/ssol/main).
From the examples it transpired that it is not just stories that spawn such emotional reactions, but much more so the formal make-up of narrative. (Which explains, by the way, why writers spend such inordinate amount of talent and time to find a formulation that fits the purpose they aim at.) The illustrations involved the results of controlled reading experiments that were set up to explore the emotional involvement provoked by these linguistic features. They were based on some premises: they started from explicitly formulated hypotheses based on predictions logically derived from theoretical insights. By confronting these hypotheses (after a statistical analysis) with independent data provided by groups of readers it was possible to find evidence in favor of the concomitant theories -but also to observe inaccuracies which may lead to revision of the theory concerned.
These illustrations have highlighted the need for such empirical methods in the pursuit of the emotional workings of literature. The methodology supersedes the meaningless search for ever new "paradigms" in which literary studies has been imprisoned over the past decades, and provides reliable, interesting and sometimes also unexpected results that can subsequently be fed back into the theories, thus guaranteeing progress in our understanding of literature, and thereby, ultimately, in understanding ourselves as human beings.