On literary meaning

The question of literary meaning is closely related to the understanding of literary interpretation, a central literary-critical concern, and very much has been written about literary interpretation and literary meaning within the discipline. I will make no attempt to review the extensive discussion.¹ Instead, my reflections in this essay have two other aims. I want to draw attention to the extent to which current writing about literary meaning within literary studies employs a perspective from inside the literary experience—what I call (loosely) a “phenomenological” perspective—and the extent to which it relies on metaphor. Further, I want to argue that it is desirable to also be able to adopt a more “naturalistic” and more literal perspective on what happens when literature is being created, disseminated, and read, and I will present such a view. My take on literary meaning will inevitably be controversial, but it should provide food for thought.

Thinking about meaning, whether literary or non-literary meaning, cannot productively be disconnected from overall ideas about the workings of language in use. I will want to emphasize the importance of this larger picture and will begin by considering three different conceptions of verbal discourse.

The ordinary-language model of verbal communication and the standard theoretical model.

As the linguist Michael Reddy has demonstrated, everyday language holds a special understanding of verbal communication in readiness for us. This is a picture of linguistic intercourse in which the speaker or writer places mental content—such things as ideas and feelings—in a kind of physical container that is, normally, made of words. On this point Reddy (1979, p. 190) refers to our use of formulations like: “When you have a good idea, try to capture it immediately in words.”

When the mental content has been captured in words, it will find itself inside the verbal container. We can say things like: “The introduction has a great deal of thought content.” And when the words reach listeners or readers these will seek to retrieve the content inside. For example: “John says that he cannot find your idea anywhere in the passage.” (Reddy 1979, pp. 191 and 193 respectively.)

Reddy lists 141 slightly different types of idiom representing aspects of this model. His examples, which are not referenced, can sound slightly artificial, but expressions following the pattern he indicates are certainly everywhere in normal talk or writing about communication. “The point of reading is the transmission of ideas from one mind to another” writes Georgina Kleege in a recent article (2020, p. 212), with a formulation fitting Reddy’s scheme perfectly. I will call the model described by Reddy “the ordinary-language model” of how verbal compositions are created, disseminated, and received.

Even though we speak and write in this fashion, few of us will actually believe that ideas can literally be moved out of people’s heads, or that words can actually contain meaning inside them, meaning which may then be extracted by receivers. Yet Reddy, whose observations about these metaphors are widely accepted by linguists, is convinced that the way of speaking and writing he describes nevertheless exerts considerable influence on our actual thinking about what happens when we communicate.

It appears psychologically easy to experience literary works on the pattern of the ordinary-language model. Take the last four lines of Louise Glück’s “The Night Migrations” (2006, p. 1). In her poem Glück reflects on what it is like to be dead, on what comforts the soul can expect when all sensual experience has ceased to exist. Her poem ends with these words:

I tell myself maybe it won’t need.

these pleasures anymore;

maybe just not being is simply enough,

hard as that is to imagine.

Trained readers will no doubt spontaneously experience the lines as if the squiggles seen and the meaning perceived were just aspects of one and the same physically present object. The idea of a verbal composition as a material entity with meaning indelibly attached to it has a kind of phenomenological validity hard to deny.² Indeed, our most basic communication models feature a sender, a message, and a receiver, and the message must clearly be the physical container of meaning described by Reddy: the message has to be a physical something (since it can be heard or read) and at the same time has to carry or contain immaterial content (since the message is supposed to be meaningful: more or less full of meaning).

Theoretical thinking about verbal communication in disciplines like linguistics and philosophy does not, of course, underwrite the ordinary-language model. Yet the understanding of the communication model’s ambiguous message slot represents a problem for serious thought. (Livingston 2021 offers an overview of the philosophical thinking about these matters.) There are various strategies for dealing with this issue, but the most common one is to separate the physical and the immaterial aspects of the message by positing the existence of an abstract object, the verbal composition itself, distinct from its physical copies. Speaking in terms of poetry: there will be the author, the reader, the physical copies, and also the poem itself. I will call this version “the standard theoretical model.”

The matter-and-mind model of verbal communication.

I have no faith in the standard theoretical model either. I find it difficult to think about verbal communication without counting with senders and receivers and physical utterances—authors, readers, physical copies—but I can manage without such objects as poems themselves. I will call my preferred picture of verbal communication “the mind-and-matter model,” for when thinking seriously about verbal communication I only reckon with minds (people’s mental contents) and physical entities (acting people considered as physical bodies; uttered sounds; physical marks in manuscripts or physical copies).

Imagine that you are telling a friend about things that happened to you yesterday. According to the mind-and-matter model there are the physical sounds coming out of your mouth, and the word-and-meaning-ideas in your own mind accompanying your talking, and the word-and-meaning-ideas arising in your friend’s mind. But if we are to believe the standard theoretical model there is also, automatically, your story itself. Your story itself is not in the physical world; it is an abstraction, but it does exist. And your story itself contains the true meaning of what you were saying. In fact, this true meaning of the story may well differ from both your own intended meaning and the meaning perceived by your friend. This is a conception that I find myself unable to accept. And the situation will be, at bottom, the same when it comes to other kinds of verbal composition, including poems and short stories and novels.

Let us return to Glück’s “The Night Migrations.” The physical marks in copies of the poem are of course meant to be read as letters, words, and sentences and to give rise to meaning-ideas and literary experiences in readers. There will have been such ideas and experiences in Glück’s mind while she was creating her poem and when she was rereading it, no doubt somewhat different ideas and experiences at different times. Likewise, there have been and will be meaning-ideas and literary experiences in the minds of readers of physical copies of her poem. These ideas and experiences will probably have a very similar content where purely verbal understanding is concerned, for “The Night Migrations” is not difficult in that respect. The deeper understanding of the poem is a different matter. And it goes without saying that the literary experiences of its readers—constituting, I want to believe, the poem’s main raison d’être—are inevitably individual, personal.

In my view, in our theoretical understanding of what happens when literature is written and read we do well to count with ideas and experiences in authors and readers, and with manuscripts and physical copies, but to forget about literary works themselves. As the philosopher Richard Rudner once expressed it: references to such things as literary works are “convenient shorthand” (1950, p. 385).³ Referring to Glück’s poem itself—as I already did here, myself, a couple of times—liberates us from taking account of the actual underlying multiplicity: the differences between copies and, above all, the differences between different human experiences of such copies. This way of expressing oneself is most often practical, “convenient”—but it also obscures important things, since it is in “shorthand”: since it represents a radical simplification.

My ideas on this point will remind many of Stanley Fish’s theses in Is There a Text in This Class?. There are indeed important similarities between Fish’s position in his book from 1980 and my own stance, but also crucial differences. Put simply, Fish held that there is no objectivity in interpretation. When readers read, their perceptions of meaning are determined by their belonging to some specific interpretive community and are the results of pure construction all the way down. For Fish, there is no literary work itself. Even physical copies are socially determined constructs, “made and not found” by the readers (p. 331).

For my own part, I have no need for the concept of an interpretive community, so vital to Fish in his 1980 book: if one wants to really understand the flexible intersubjectivity that certainly exists in reading, this will call for a much more nuanced approach. Even more importantly, I regard the physical matter that we conceptualize as a physical copy of a literary work as being actually there, independently of human perceptions. The physical marks meant to be read as letters et cetera are, to me, physically real, and the physical marks in a copy of the poem by Glück will differ from the physical marks in the copy of any previous poem. For this reason competent readers will be in a position to construe her poem in its own unique way, linguistically and also with respect to literary meaning.⁴

How do meanings come into being?

At this point many will protest that literary works do exist, so that there is no way around them in a theory of literature, and that literary works have meaning. My answer is that we should not trust received opinion and so-called common sense without critical reflection. It is a metaphysical question what truly exists, what reality is truly like. What I am after, here, is as sensible and credible an account as possible of how language works. My unwillingness to endorse the standard theoretical model (and, of course, the ordinary-language one) is motivated by my conviction that the model does not hang together intellectually.

The question of meaning is key to my scepticism toward verbal compositions as separate genuinely existing abstract entities. If one adheres to the standard theoretical model, one will believe that Glück’s “The Night Migrations” truly exists in our world as an abstract object, not accessible to the senses and without any physical location. One will also believe that this abstract object has some body of meaning associated with it (also immaterial, of course). One may believe that the body of meaning is fixed once and for all, or that it can change over time. One may believe that the body of meaning is, at any specific time, determinate, or that it is instead indeterminate, perhaps even in flux. But no matter whether one is a traditionalist or a poststructuralist, one will believe that there is, somehow, some body of meaning associated with “The Night Migrations” at any given time. One will not believe that Glück’s poem is meaningless. It is this belief in objectively existing bodies of meaning in connection with literary works that I find impossible to motivate. Glück had her meaning-ideas; her readers will have theirs.

It is widely thought that language, or linguistic conventions, supply sounds and marks with meaning. Philosophers sometimes suppose that conventions—for example, “the linguistic, cultural, and artistic conventions operative at the time the work was produced”—can give linguistic expressions a more or less determinate meaning. (Gaskin 2013, pp. 217–219—quote from p. 219—and see also, for a much earlier expression of this kind of idea, Beardsley 1981 [1958], p. 25.) And in literary studies it was, for decades (and perhaps still is), common to suppose that language creates meaning but not determinate meaning, that “it is characteristic of language to generate infinite webs of meaning, so that all texts are necessarily self-contradictory” (Barry 2017, p. 37).

However, unlike physical utterances in speech or writing, language itself and linguistic conventions are abstractions. Abstractions do not, by definition, belong to the physical world and they are not, therefore, supposed to be able to cause anything at all, neither anything concrete nor anything abstract.⁵ Human beings with literacy and relevant linguistic knowledge can see and process marks intended to function as letters, and they can construe meaning-ideas on this basis. How language or linguistic conventions could bring meaning into being without human intervention appears inexplicable to me, no matter what meaning is supposed to consist in more exactly.

It will be of no avail to look to professional linguistics for an explanation. Contemporary linguistics typically identifies the meaning of a verbal composition, its so-called “utterance meaning,” with what the creator of the composition meant, the so-called “speaker meaning.” Sentences (sentences in the abstract, so-called sentence-types) are supposed to possess sentence meanings. When used by a speaker or writer, the sentence is said to carry this sentence meaning, very often somewhat modified by its appearance in its occasional context. On the basis of sentence meaning and context one is expected to be able to draw a fair conclusion about what the sender meant, provided that the sender was using the language correctly (cf., e.g., Cruse 2011, pp. 24–26 and Birner 2013, p. 24).

The same problem emerges here. Meaning (whatever meaning is supposed to be—linguistics has no clear answer to offer⁶) cannot very well be produced, in something like a quasi-chemical reaction, by abstract sentence-types and concrete context working on the physical sounds or marks produced by the sender. The linguist will have to come in and determine the reasonable interpretation of the utterance, weighing all relevant factors together. What we will have is, again, meaning-ideas: not those of the sender or the receiver but those of a third party, the linguist.

In Jacques Derrida’s thinking about language—for a long time highly influential within literary studies—it was a central idea that physical marks meant to function as linguistic signs can always be read and reread: for him, to write “is to produce a mark that will constitute a sort of machine which is productive in turn, and which my future disappearance will not, in principle, hinder in its functioning, offering things and itself to be read and to be rewritten” (1988a, p. 8). Derrida, who did not distinguish clearly between physical marks and linguistic expressions, held that such uttered linguistic expressions are iterable. He emphatically denied that he simply referred, by this, to the trivial fact that one can use the same linguistic expression again and again (see esp. Derrida 1988b, pp. 119 and 127). I believe that his idea was, rather, more or less this: one and the same uttered linguistic expression bears within itself, as a possibility, meanings it can have in the possible contexts in which it can appear.⁷ For this reason verbal utterance will always display a radical form of indeterminacy. When one creates a written mark, the mark will, so to speak, carry within itself the possibility of a multiplicity of meanings.

This is not how I think about these matters. Physical marks meant to function as a sign can be read and can certainly, potentially, give rise to different meaning-ideas. This does not mean that the different meaning-ideas are somehow inherent in the physical marks themselves, forming part of the meaning of the sign, so that a verbal composition objectively contains disparate meanings.⁸ Physical marks in themselves are just that, physical marks, nothing more.

In short, I can find no reasonable explanation of how abstract objective bodies of meaning associated with abstract objective verbal compositions could possibly come into being. But I have nothing against saying that people form ideas in their minds, so I am ready to hold that Glück formed meaning-ideas in her mind when preparing the manuscript of “Night Migrations,” and that I, similarly, form meaning-ideas in my mind when I read the quoted lines on the first page of a copy of her collection Averno, and that literary critics writing about the poem can form their own individual meaning-ideas. True, I cannot give an account in scientific terms—for example, in neurological terms—of what concretely happens when ideas with a given content are formed in the mind (nobody can), but I know of no better way of conceiving of these matters. The idea that language creates meaning does not make sense to me, nor the ideas that conventions create meaning or that physical marks can be meaning-producing machines.

It is easy to reply that such conceptions may not make sense to me, but that they will make sense to many other people. Of course. The matters brought up here are deeply divisive and my brief discussion has perhaps achieved nothing more than to point to important areas of dissent.

On the non-objectivity of literary meaning.

One objection to my reasoning about meaning lies near at hand: must there not be at least some objective meaning in connection with literary works, since it is apparently possible to say true things about what they mean? For is it not true that the lines from Glück mean that the speaker tells herself that it is hard to imagine that the soul won’t need, after death, the pleasures that we get through the senses, that simply not being will be enough for it? In that case, are we not faced with an example of objective meaning in connection with a literary work?

“Objective” can mean at least two different things—the word can mean “unbiased” or it can mean “in the objects themselves.” The italicized description of the meaning of Glück’s lines is objective in the first sense: I feel certain that every competent reader will largely agree with this suggestion of what her lines mean. There will then be consensus among competent readers about this understanding of Glück’s lines: the meaning-ideas of competent readers will coincide in this respect. Whether you also want to call their meaning-ideas true will depend on how you wish to use the not uncomplicated concept of truth. (I do not mind to call the ideas true, in a casual sense of “true,” but if the theory of truth is seriously at issue I will take care to distinguish between consensus and truth.) But, in any case, there will be no objective meaning in the second sense. If you look at a physical copy of Glück’s “The Night Migrations”—and, remember, this is the best you can do: you can never look at “The Night Migrations” itself—there is no objective meaning in the physical object itself, neither in the white paper nor in the pattern of printer’s ink (nor in the pixels, if you read the poem in electronic form). A meaning—that is, meaning-ideas—has to be construed on the basis of what you can see and read. The physical object does not have meaning; it has meaning ascribed to it. Different readers can ascribe more or less the same meaning. Then one can call the meaning “objective,” in the first sense, but the meaning will still not be in the thing itself.

This is a crucial point. People in general, but also linguists and philosophers, like to say that we can utter words, write down sentences, et cetera. Words are supposed to have word-meanings and sentences sentence-meanings. If the sounds we utter and the marks we write down when communicating verbally are really, already in their capacity of physical sound and physical marks, words and sentences with a meaning, then it seems that there is really objective meaning in the second sense, meaning located in the physical sounds or marks themselves.

However, in my view, the talk of uttering words et cetera should not be taken at face value, just as little as the talk of transmitting ideas between minds. Everyday language is there to be practical but not necessarily logical. It should be obvious that receivers need to interpret the sounds made in speech or the marks produced in writing if they are to turn these into linguistic expressions. One may frown at the word “interpret,” for what happens is mostly something done without any conscious attention at all. What I want to press home is the fact that mental processing is needed already to turn the visual impression of my quotation from Glück into a series of English words. The word “I” is not physically there on the page; a physical mark is there. The mark is undoubtedly meant to be read as representing the Roman capital letter ˂I˃ and as representing, thereby, the English pronoun “I”; this is also how the mark will be understood by competent readers. But these facts do not make the physical mark itself into a word.

It is worth emphasizing that what I just discussed was the absolutely basic, purely verbal, meaning of Glück’s poem. Literary meaning is something much more inclusive, comprising also the poem’s “deeper” meaning, no matter how one wishes to develop the idea of its literary substance. Consensus will be far less likely when readers’ individual literary experiences are taken into consideration. Very many empirical studies, from I.A. Richards’ Practical Criticism onward, make this fact impossible to neglect. (Richards 1929; two later book-length studies of different readers reading the same work or works are Holland 1975 and Pette 2001.)

Since it is commonly taken for granted that linguistic compositions, if in any way normal, have bodies of meaning indelibly associated with them, few people have wanted to formally argue for this idea. Still I know of certain arguments, and among these I find it motivated to comment briefly on considerations by the philosophers Jerrold Levinson and Paisley Livingston.

Levinson’s reasons for thinking that statements about the meaning of artworks can be true build on negative characterizations like “Stevens’s The Emperor of Ice Cream is not about the frozen dessert industry” (1992, p. 240). In this case too my answer is to refer to consensus—here, negative consensus. No competent reader would understand Wallace Stevens’ poem “The Emperor of Ice Cream” to have the frozen dessert industry as its theme. This does not prove that the poem has a true meaning.

Livingston, for his part, has argued that there must be a “fact of the matter” about what such things as literary works mean (2005, p. 144). He takes it for granted that non-literary works have determinate meanings, and holds that we cannot draw a sharp boundary between literary and non-literary works. My reply to this is that, contrary to what is commonly thought, non-literary works do not have objective meanings in the full sense either. There is no meaning hidden in the printer’s ink of a non-fiction book, but there can of course be interpretive consensus among its readers.

Livingston also introduces the idea that we could not agree or disagree about the meaning of a work if there were no fact of the matter (2005, p. 144). Again, this fails to convince me. People can be under the false impression that there is a truth about the meaning of the work. And, more importantly, we can agree or disagree, and offer rational arguments, also when we know that there is no truth about what is right. We can do so about many important or trifling matters in life: about what political party one should vote for, about what to have for dinner, and so forth.

I have explained why I do not believe that objective meaning, in the sense of meaning inherent in the utterances themselves, can ever come into being, and arguments to the contrary—in so far as I am familiar with such arguments—do not seem to me to have any force. I will now switch track: I will argue that problems of the kind just discussed are highly relevant for literary studies. The next section will be devoted to a look at how literary specialists today tend to express themselves about literary works and their meanings.

Critical discourse about literary meaning.

At the opening of my essay I claimed that contemporary critical discourse about literary works and their meanings is often metaphorical and often formulated from a “phenomenological” perspective. In what follows I will concentrate exclusively on these two features.

The literary work or text is practically always being referred to as a genuine object. (But see also note 9 below.) It is not necessarily clear what kind of object, but the metaphorical character of formulations along these lines becomes obvious as soon as one scrapes the surface. Sometimes the work is described as something decidedly physical, much in the fashion of the ordinary-language model; for example: “A text is not an immaterial entity but a tangible thing whose very materiality affects its meaning” (Andersen 2017, p. 311). Yet if a work like “The Night Migrations” is a tangible thing it must obviously have a definite location in the physical world, and one may ask where exactly one can find it and place one’s hands on it. Conversely, the work is sometimes conceived as something immaterial that is, so to speak, anchored in the physical world by means of its concrete medium, which constitutes “the haptic, phenomenological interface between a work and its apprehension by a subject” (Hayot 2012, p. 177). This too, must be metaphorical: a physical copy and an immaterial work cannot literally interact or share a common boundary.

Material or immaterial, the work is typically presented as having an extension in space, in particular, as having surface and depth. Thus literary works can be thought to “contain a deeper dimension worthy of interpretation, a dimension which it is important to bring to the surface,” and interpretation and explanation can be expected to “unfold, as it were, the meaning packed into the text” (Selbmann 2002, p. 37; my translation). Remarks in this vein must also be de facto metaphorical. Nobody really believes that meaning of any kind is physical in nature so that it can have an extension in space and, literally, surface and depth. And if meaning is immaterial it is not in any space at all. This, by the way, is one of the problems with Roman Ingarden’s classical, influential theory of what a literary work is. According to his phenomenological analysis—phenomenological in the true philosophical sense—the work is not of a physical nature, but it is still said to consist of a series of strata: linguistic sound formations, meaning units, represented objects, schematized aspects. These layers seem to form a spatial structure housed in a non-spatial space—for Ingarden: a kind of non-physical intersubjective space. (See Ingarden 1973, part II; about the intersubjective character of the literary work see esp. § 66.)

Literary works, then, are typically understood as objects.⁹ These objects are often depicted not as passive things but as possessing agency in their own right. Referring to verbal compositions as active agents is common in verbal discourse: in an earlier section I myself wrote that “my brief discussion has perhaps achieved nothing more than to point to important areas of dissent,” as if my discussion could achieve something on its own, entirely independently. This is a type of innocent figure that can be encountered everywhere language is being used. However, the image of the literary work (or the literary text) as an agent often becomes more pronounced, even radical, in critical discourse.

A case in point is the very common image of the reading of literature as a dialogue with the work. Work and reader can be said to “interact,” so that reading becomes a “two-way transaction” (Felski 2008, pp. 11 and 18 respectively), or the reading of literature can be said to take the form of “a conversation between the work and the reader” (Moi 2017, p. 217). This way of writing, too, metaphorically supplies the literary work with a mind, but it does so in a more expansive manner: if able to literally enter into a dialogue, the literary work must not only have a consciousness but also the competence needed to communicate.

In the intercourse between a literary work and a reader, the literary work is, typically, understood to be at least on an equal footing. Indeed, it may be suggested that the best you can do as a reader is to simply let the literary work act on you. Wolfgang Iser once wrote that “the reader comes closest to the text game when the awareness of being played by the text grows” (Iser 1993, p. 275).

I noted, earlier, that it is easy to experience physical lines like the ones from Glück as a material object somehow embodying meaning. The metaphors brought up in this section appear to me to be, likewise, grounded in the spontaneous experience of reading. The meaning springing up in our minds when we read seems to be something we encounter and grapple with, as if it comes from a source. If one wants to keep the author out of the picture—as literary critics often want to do—it can seem that this source will have to be the literary work. The more attention one pays to these meanings, the more details or aspects may come to one’s mind, as if one penetrated deeper and deeper into a space of meaning difficult to survey. Consequently, when reading, and not least when reading literature, it is not unnatural to be aware of one’s own activity, and at the same time of the difficulty of doing justice to all the impressions arising. When new perspectives and possibilities constantly emerge, it can feel as if the things read resist one’s full understanding. Then it can appear entirely justified to describe the work as something actually acting on one, or as a partner in a communicative exchange.

It is worth emphasizing that the quotations in this section all come from theoretical articles or books. The fundamental reliance on metaphor and on the phenomenological perspective can easily be found in serious theorizing within the discipline. Two more examples.

Referring to what things seem like “in one’s experience of interpretation” Jonathan Culler has maintained that “meaning is both the semantic effects one experiences and a property of the text against which one seeks to check one’s experience” (Culler 1983, p. 132; he adds the caveat that “our experience is scarcely a reliable guide in these matters”). In a later formulation: meaning is “simultaneously an experience of a subject and a property of a text” (Culler [1997] 2011, p. 68). In this version of what meaning is, the literary work and the reader face each other like two equals, and meaning is said to reside in both. But whatever meaning is considered to be, can it really be found in two places at once? Or how is Culler to be understood in more literal terms? Later he can seem to give the literary work the upper hand and, using the metaphor of the agency of the literary work, sympathize with the idea that the meaning is what works “achieve,” what a text “ends up doing” (Culler 2007, p. 225).

Derek Attridge, for his part, distinguishes between a verbal text—described in the spirit of the ordinary-language model as “the set of verbal signs embodied in a particular object”—and the literary work that may emerge if the text and the reader permits: a literary work is “a realization of the text as it is experienced in my reading” (Attridge 2015, pp. 34 and 30 respectively, italics in the original). This can seem to mean that a literary work—for instance, “The Night Migrations”—must be as many literary works as there are readers, or readings.¹⁰ Yet Attridge maintains—mobilizing a different conception—that literary works exist “neither wholly objectively in the world nor wholly subjectively in the mind of the reader, but in the relation between them” (2015, p. 90). Here the work is not identical with what is experienced in the reader’s reading but exists in the relation between the reader and what is (very broadly) called “the world.” The phenomenological character of Attridge’s analysis, its nature of an attempt to do justice to a lived experience, appears obvious to me, and Attridge himself writes that he is looking for “the essence of the literary experience” and defends his dependence on introspection (2015, p. 90 and, concerning introspection, p. 93).

The “phenomenological” perspective and the “naturalistic” one.

The “phenomenological” way of viewing things is not the only possible one. As I suggested at the very beginning there is also what could be called the “naturalistic” perspective. Think of the rainbow phenomenon. Seeing a rainbow affords a visual experience, a phenomenological access to rainbows. Yet there is of course also the naturalistic perspective: the physical explanation in terms of sunlight refracted by raindrops. One can speak, similarly, of a perspective on what happens when literature is written and read that does not take the reader’s literary experience as its point of departure. Of the three models discussed earlier, at least my matter-and-mind model is, in that sense, “naturalistic.”

Literature is there to be read and experienced. It goes without saying that one has to adopt a phenomenological perspective if one is to get a literary experience out of a poem like Glück’s. A reader who is reading for the literary experience, and not for making literary-critical observations, can be expected to read and establish a purely verbal meaning. This purely verbal meaning will give rise to further perceptions, thoughts, and feelings—normally largely implicit—and all this will form the content of the reader’s literary experience, which, if one wishes, can be called the literary meaning of the poem (for this reader, in this reading). How readers who read for the literary experience actually read and experience in more detail is a question for empirical literary aesthetics which I will not try to address here. (For empirical perspectives see, e.g., Miall 2006 and Winner 2018.) The role of the phenomenological perspective in literary criticism is, for me, more ambiguous; I will come to this in a moment.

In the previous section there was also much about the metaphorical character of critical discourse about texts and meanings, and those who sensed reservation or opposition are not entirely wrong. I am not an enemy of metaphor, and it is possible to understand much of the figurative language that I commented on as harmless and as lending life and colour to literary criticism. Indeed, these metaphors are so much a part of everyday language that they can be avoided only at the cost of some pedantry and unnaturalness. Yet if one seeks a naturalistic account of what happens when literature is written and read, metaphors will have to be kept out as much as possible. One will be looking for the best possible explanation, not for a metaphorical one.

This is not necessarily how this matter is viewed within the discipline. Some would hold that some of the notions I call metaphorical actually represent a reasonable way of viewing things. This is so, for example, with respect to the idea that literary works can genuinely perform actions. (Culler 2007, p. 128–129 offers an example; see my discussion of his argument in Pettersson 2017, p. 139.) Yet viewing a work as a genuine agent must imply viewing the work as equipped with a mind, which to me seems obviously wrong. It is worth noting that references to actor-network theory (as in Felski 2020, esp. pp. 64 and 140) work in a similar direction by reinterpreting agency: every factor relevant to a process will be seen as an “actor.”

Admittedly, there are those who do not believe in the kind of rationality on which I rely here. In a book from 2003 (p. xvi) Stathis Gourgouris explicitly endorsed the idea that literature can think; when doing so he adopted “a perspective that jettisons the constraints of the truth/falsity framework.” Similarly, in a sympathizing discussion of the postmodern, Andrew Bennett and Nicholas Royle draw attention to its emphasis on how the law of non-contradiction “may be productively questioned or suspended,” and to how postmodernism has suspended and deconstructed “the opposition between the rational and the irrational” (2016, pp. 326 and 327 respectively). At points like these there opens a divide that makes further discussion difficult.

According to a less radical idea literature is not fully amenable to rational explanation. In Attridge’s words: “The power of literature may well lie primarily in what can’t be scientifically explained.” (2015, p. 100) With this I entirely agree. I think of the literary experience as being at the heart of what literature is about, and its contents and mechanisms are certainly not fully within the reach of science and may never be (which can be said, of course, of nearly every kind of human experience). Still I do believe that many relevant and useful things can be said, from a naturalistic perspective, about what happens when literature is written and read. I find discussions like the ones at the opening of my essay clarifying.

But are such discussions really relevant for literary studies? What good do they do for literary criticism, what new angles do they add? How can they affect our critical projects? Questions of this kind are natural to ask if one views the interpretation of literary works as the core of the discipline and the introduction of new ways of approaching literary works, adding something to our interpretive arsenal, as the only valid motivation for literary theory. For me, however, literary studies is about understanding the phenomenon of literature, which potentially opens for a huge array of historical, interpretive, and theoretical approaches. To attempt to comprehend, as well as one can, what happens when literature is written and read, seems to me to be to be a relevant thing for the literary specialist to do, irrespective of the role of such comprehension for the interpretation of literary works. It has a value in itself.¹¹

Literary criticism and literary meaning.

It is certainly true that comprehension of what happens when literature is written and read does not in itself offer a critical programme. Understanding what a situation is like never automatically tells you what to do, whether it is a question of choosing a political party, a promising dinner, or a critical project. Yet it seems to me that literary critics need to be able to adopt both the phenomenological and the naturalistic perspective. They must obviously be able to read and experience literary works in the manner of readers for the literary experience. But it is arguably also crucial for literary critics to have as literal an understanding as possible of what happens when literature is being written and read. For example, belief in the existence of a body of meaning indelibly associated with a literary work (and not “just” a larger or smaller domain of consensus in people’s meaning-ideas) can give rise to the not uncommon idea that different critical approaches are in competition for the right way of grasping the meaning of works (and not “just” asking different research questions on the basis of different frames of interest).¹² Critics who see clearly that literary criticism is not simply an attempt to approach the postulated meaning of the literary work in question will feel the pressure to explain to themselves what their critical projects are aiming at more precisely, and to bring their more exact purpose, and its special logic, to conscious attention.

When it comes to literary interpretation critics can pursue a variety of different goals. A critic writing on “The Night Migrations” may want to describe it as forming part of a kind of poetic project of Glück’s; then it will become crucial to try to understand what Glück herself may have wanted to achieve and how she herself may have conceived of the poem. Or the critic may put the theme of death and afterlife centre stage and attempt to say interesting things about how this theme appears to us, or can be made to appear to us, in the poem, perhaps opening comparative perspectives while doing so. Or the critic may want to present an innovative and rewarding reading of some other kind, attempting to push readers’ ways of perceiving the poem in new directions.

The options are endless; the critic can pursue so many possible goals. If one does not believe that there are true, objective meanings, determinate or indeterminate, associated with what we call literary works, then a key question to an interpretation will be what its point is supposed to be, what question the critic is attempting to answer or what other purpose the critic is attempting to make the interpretation fulfil. It will be too simple—it will be to misunderstand the situation—to just point to the understanding of the literary meaning of the work, or of what the work is “doing,” as one’s aim.

Nor will critics who do not believe in objective bodies of meaning associated with literary works be under the even more widespread impression that literary critics, as the professionals, are the real experts when it comes to understanding literary meaning, and hence literature, while lay readers are more or less negligible amateurs. It is a recurrent theme in critical discourse that literary critics are doing the same thing as ordinary readers—namely, reading literature—but that they are doing this better. (See, e.g., Guillory 2000, pp. 31–32, and Auyoung 2020, esp. p. 103.) The truth, as I see it, is that reading literature for the literary experience is one thing, commenting on it, from some of the innumerable possible points of departure offering themselves, is another. Critics can certainly read literature for the literary experience, but when they act in their capacity of critics they are commenting on literature, not reading literature. It should be a sobering thought that Glück hardly wrote her poem mainly as a means of offering material for critical interpretation or analysis. Literary criticism is an important pursuit in its own right, but “ordinary” reading, reading for the literary experience, is arguably what literature is for. Belief in objective literary meaning, and in expertise about it, tends to lead to misconceptions regarding literary criticism, ordinary reading, and the art of literature itself.