Patrick Colm Hogan, University of Connecticut

Recently, I read a number of Seventeenth-Century Chinese stories as well as some classical Roman works that reminded me of one another and that led me to recall some South Asian and Native American stories that I had read earlier. They were all versions of the cross-culturally recurring romantic story, but they were not prototypical. The interesting thing was that their deviations from the prototype were closely related to one another, clustering into a sub-genre. As I reflected more on this sub-genre, I realized that the pattern extended quite widely. The present essay is an attempt to define that sub-genre, to sketch some of its constituent varieties, and to suggest some possible explanations for both. More precisely, I refer to the sub-genre as “intensified romance.” It is intensified in the sense that it makes the lovers’ union appear to be not only socially problematic, but in principle impossible. That impossibility may be either ontological or moral, yielding two types of the sub-genre.

As some readers of this essay were concerned about how I might explain the cross-cultural recurrence of love stories more broadly, I should note that the present essay is not aiming to explain the genre in general. I have discussed that topic in The Mind and Its Stories, Affective Narratology, and elsewhere (for a summary of the account, see “Story”). The present essay explores one specific variant of the romantic story and seeks to give a preliminary account of that variant alone.

A Note on Simulation

Simulation is a cognitive process by which we imagine trajectories of actions or events (for an introduction to simulation and its relation to literature, see Hogan How; see also Oatley, “Meetings”). That process may be a matter of interpolating causal links between real events that we experience, projecting future outcomes of current conditions, envisioning counterfactual alternatives to what has occurred, or creating purely hypothetical scenarios. A key evolutionary function of simulation is to allow us to consider alternative courses of action “offline,” thus without the consequences of real activity. For example, if I simulate a dangerous course of action—say, swimming at a beach closed due to shark sightings—I incur no risk and may decide to avoid real swimming on the basis of the simulation (including its frightful image of the malevolent leviathan crushing my thighbones like toothpicks between its teeth). If I could learn only from actual experience, I might end up eaten by a shark the first time I am faced with such a situation.

For simulation to fulfill its evolutionary function, we must be able to make adjustments to our simulations at various levels. For example, suppose an office worker wants a raise in her salary. She has to be able to consider the various, broad strategies for pursuing this—applying for other jobs, arguing for particular salary evaluation procedures, engaging in a class action suit, approaching her supervisor to discuss an individual increment, and so on. In each case, she must be able to imagine more specific strategies as well—what would be the best evaluation procedures or just what points she should make to the supervisor in support of an individual increment. In connection with this, we may draw a relative distinction between two sorts of variables—first, parameters, which define classes of alternatives; second, local variables, which specify instances of those alternatives.

Stories are the product of simulation. Authors create stories by imagining trajectories of actions and events, commonly through synthesizing genre prototypes with memories of actual experiences, hearsay, trajectories from other stories, and other sources. Part of this simulation involves making both larger and smaller changes, thus altering parameters to produce types of story and altering local variables to produce particular stories.

In The Mind and Its Stories and Affective Narratology, I have argued that perhaps the most common cross-cultural genre or story universal is romantic tragi-comedy. In its most prototypical form, this genre involves two people who fall in love, but are prevented from uniting, often by their parents and commonly due to some social conflict (such as the opposition between the Montagues and Capulets in Romeo and Juliet). They are separated, frequently with suggestions or imagery of death. A rival may enter, producing a love triangle, before the lovers are finally united (or, in the tragic version, prevented forever from uniting).

The preceding comments on the nature of simulation would lead us to expect that there will be versions of any given genre that are produced by parametric variation. These are presumably most likely—as most consequential—at story “junctures,” the points at which significant causal changes occur in the trajectory of actions and events (on junctures, see my The Mind 91-93). One such juncture is governed by the nature of the restriction that separates the lovers. In most of the cases I considered in earlier publications, that restriction is contingent and social. Again, the Montague-Capulet feud comes to mind. There is no necessary reason why these two families must be enemies or why their children could not be united. However, there are other possible types of restriction. Specifically, some restrictions may be felt to be necessary, either necessary in the ethical sense or necessary in the ontological sense. In other words, the lovers’ union may be experienced as in some way unthinkable or impossible, either morally or practically.

The clearest case of moral intensification concerns incest. Ontological intensification bears most often on the union of humans and spirits of some sort—either superhuman spirits (such as gods, angels, or ghosts) or (usually subhuman) animal spirits. I will refer to the superhuman/human/subhuman order as ontological or, drawing on usage in linguistics, as manifesting an animacy hierarchy. Of course, it is not strictly correct to say that these are literally unthinkable or impossible. The moral violations occur in the real world, though they are paradigmatic of sexual taboo. The ontological violations are possible in the storyworlds of some fictions, though (presumably) not in the real world, unless the ontological difference is interpreted metaphorically. The point is that, in each case, there is something about the nature of the lovers that, in certain circumstances, makes it appear impossible for them to be united (in the sense that, on hearing of such a union in real life, one might exclaim, “That can’t be!,” as in “A mother marrying her own son? That can’t be!”). We may refer to the broad varieties of romance—where the restrictions on the lovers are contingent and social, on the one hand, versus necessary, on the other hand—as basic and intensified respectively. We may further subdivide the intensified varieties into moral and ontological.

We will consider some cases of each sort of intensified romance in the following sections. Before going on to that, however, it is useful to note that another variable or set of variables enters here. This concerns just who knows about the problematic identity of the lovers and just when they know. Readers of Romeo and Juliet may recall that the titular lovers become enamored of one another before learning that they are from the opposed families. Sometimes, the ethical and ontological versions of the romantic plot involve the lovers knowing about their difference early on; sometimes, both lovers learn only later; sometimes, one knows while the other does not.

Finally, I should note that two motifs recur with some frequency in intensified romances. As I am using the term, motifs are sequences that can be adapted to different genres and integrated into a range of stories. (The key point is that they cross genres.) Here, the motifs in question are the concealment of identity, where a character hides some important aspect of who he or she is, and mistaken identity, where a character misattributes a category to another person. In the ethical and ontological romantic plots, the concealment and mistake most often concern one or both of the lovers.

Romantic Stories and Animacy Hierarchies (I): Humans and Animal Spirits

A story told by the Arawak (see Lévi-Strauss 256-257 and Roth 203-204) concerns a hunter and a jaguar (or tiger). The jaguar has taken the form of a human woman and joins the hunter as his wife. However, they conceal her identity. The two live together happily; she is a good wife and hunting partner. At a certain point, they visit the man’s mother. Though the jaguar woman stresses that he should never reveal her identity, the man ends up informing his mother, who conveys the information to the society at large. The jaguar woman suffers from shame and leaves her husband. Despite great effort, he is never able to find her.

This is, to my mind, an intriguing and affecting story. It also manifests a structure that recurs with variations in different traditions. A forbidden—putatively impossible—union occurs. Since it is forbidden, the lovers conceal their identities. Despite being warned not to do so, one spouse engages in some action—constituting a tragic error or failure—that leads to the exposure of the identities. This in turn forces the separation of the lovers in a tragic ending. (It may be that tragic outcomes are more common in these intensified variations on the romantic genre than in the basic form.) Note that “tragedy” and “comedy” here are what we might refer to as “modes of resolution” for the genres defined by happiness goals. For example, romantic stories may be resolved by the lovers’ achievement of enduring union (comedy) or by a sequence of events (e.g., one lover’s death) that renders such achievement impossible (tragedy). Like genres, tragedy and comedy may themselves have recurring properties (such as the commission of a tragic error).

One striking feature of the Arawak story is that the inter-species union is, in fact, not impossible. Its putative impossibility is a function of knowledge. It is only when some people know the identities of the lovers—when society learns that they are not man and woman, but man and jaguar (or tiger)—that the impossibility appears. In other words, it is not an impossibility of the lovers or categories themselves, but of social opinion about those categories. This suggests that the restriction is ultimately a social restriction, though it is represented as ontological. In keeping with this, Alexandra Aikhenvald’s research indicates that “jaguar” might suggest a low status group (12), a point consistent with the woman’s feeling of shame. Alternatively, Roth’s work may point to a link between tigers and antagonistic out-groups (367). Thus, we have the usual sorts of social taboo—due either to social hierarchy within the society or conflict with an enemy group. The ontological romance intensifies this ordinary social problem through the animacy hierarchy (hence my name for this variant, “intensified” romance).

Pu Songling’s Strange Tales, an important, late Seventeenth-Century, Chinese work, includes several stories of this sort. In “The Girl in Green,” a woman arrives at the male protagonist’s lodgings unexpectedly. Though she conceals her background from him, they have sexual relations. The man asks her to sing. She refuses, warning that it may be dangerous. Eventually, she gives in to his pleas. When she departs, she is attacked, and he has to rescue her. At this time, he learns that she is not a woman, but a hornet. She thanks him for rescuing her, but leaves forever.

Here we have the same general structure. There is an inter-species union, though one in which the man is ignorant that the woman is not really a woman, but an insect. She warns him that they should behave in a particular way; he insists otherwise. The result is the disclosure of their species difference and their consequent separation. (As is undoubtedly clear, here and below I engage in the sort of structural mapping of story sequences that Lévi-Strauss developed so powerfully in his four-volume Introduction to the Science of Mythology. As is probably no less clear, however, the conclusions of my analysis and the psychological principles it involves are quite different from those of Lévi-Strauss.)

Another story along these lines is Pu’s “Princess Lotus.” In a dream, the protagonist meets a king, who introduces him to a young, beautiful woman. The protagonist falls in love and marries the woman. Everything seems to be proceeding with remarkable calmness. But suddenly the protagonist’s in-laws are threatened with destruction. He eventually learns that they are, in fact, not people but bees. This varies the structure, since the revelation is not the result of a violated taboo. It rather involves a danger to the family of one of the lovers. But there is still inter-species union connected with identity concealment, followed by disclosure.

Staying within the Chinese tradition, we find a further instance of this story structure in the often-retold “Legend of the White Snake” (for a summary of one version, see “Legend”). A magically powerful white snake transforms herself into a woman and marries a human. When the human discovers her identity, he dies. Through supernatural means, she manages to revive him; however, further tribulations follow. In some versions, she overcomes subsequent obstacles and the lovers are reunited enduringly; in other versions, she is unsuccessful.

Romantic Stories and Animacy Hierarchies (II): Humans and Immortal Spirits

The human-immortal romances are, on the whole, more insightful and engaging than the human-animal romances. They are also more abundant, though admittedly a number of stories could be put into either category. Another story from Pu follows the pattern of “The Girl in Green” and the Arawak tale. In “Sunset,” an unknown woman arrives at the protagonist’s home. Despite the intimate relationship that develops between them, she conceals her background and prohibits questions about it. He nevertheless tries to ascertain her identity, then continues this investigation despite further warnings. The man’s family has a crisis when his niece is possessed by a spirit. The mysterious woman helps through magical intervention. In consequence, her father learns about her affair with the human and forces their separation. Thus, we have the usual separation of lovers by the parent, but in this case due to an ontological discrepancy in the identities of the lovers. The woman, it turns out, is daughter to the Golden Dragon King. They are never reunited.

This story is particularly well developed. It takes up the motif of the threat to the family of one lover, which we saw in “Princess Lotus.” It also includes concealed identity, one lover’s attempt to prevent the disclosure of that identity, and the other lover’s violation of a warning.

A comic version of human-immortal romance may be found in Li Hao-ku’s early Thirteenth-Century, Chinese drama, Chang Boils the Sea. In that play, a scholar falls in love with the daughter of a dragon king. Through the use of magic, the scholar is able to join his beloved—although she is “thirty thousand fathoms beneath the blue sea” (170; the physical inaccessibility concretizes the ontological discrepancy between human and spirit). He is also able to secure the consent of the dragon king. Of course, this is not a concealment-revelation-loss sequence. Rather, this story indicates that the human-immortal romance may follow the more routine or basic form, with parental disapproval and the lovers’ work to overcome that disapproval.

The intensified romance is not confined to Chinese and Arawak traditions. In the Indian (South Asian) tradition, there are several versions of the romance between the mortal Purūravas and the celestial nymph Urvaśī. Consider, for example, the important early collection of stories from Hindu tradition, the Viṣṇu Purāṇa (see Wilson iv-viii). In that work, Urvaśī is condemned to a period of earthly existence, where she falls in love with Purūravas. They marry on the condition that Purūravas never violate a particular taboo. He is tricked into doing so and the lovers are separated, though they are ultimately reunited when Purūravas is elevated to the status of an immortal. In Vikramōrvaśīyam, the version by the most renowned Sanskrit playwright, Kālidāsa (fifth century C.E.), the human king Purūravas and Urvaśī fall in love. They are separated, as Urvaśī lives in heaven. However, due to her preoccupation with Purūravas, she commits a serious breach of etiquette and is condemned to become mortal for a certain time. The change of Urvaśī from spirit to human allows her union with Purūravas. But the temporary nature of her incarnation limits the duration of that union. In the end, she is allowed to remain human as long as Purūravas is alive.

In these cases too we see a variation on the intensified romance. In neither version is it a matter of concealment and exposure. However, the ontological contradiction of love between a mortal and a spirit is clear. Moreover, in theViṣṇu Purāṇaversion, there is a taboo on one of the lovers and the violation of that taboo leads to their separation. Interestingly, the human-spirit union is enabled by the change of status—the fall ofUrvaśī in bothversions and the elevation of Purūravas in the Viṣṇu Purāṇa. It is also worth noting that, the play seems particularly to recall death and grief. It thereby suggests a connection between separation in the ontological romance and the more mundane issue of bereavement, where the (still living) human lover is separated from (his or her deceased) beloved, who is now a spirit.

A particularly extended and complex version of the structure may be found in the Roman story of Amor and Psyche, as developed by Apuleius in the late second century. Due to a curse and subsequent abandonment by her family, the beautiful (human) Psyche ends up married to (immortal) Amor. However, Amor conceals his identity, and even prevents Psyche from seeing him. In keeping with the pattern we have considered, he warns Psyche that she should not seek to discover who he is. Convinced by her (jealous) sisters, Psyche tries to uncover her spouse’s identity, planning in fact to kill him. Instead of the dangerous monster she anticipates, Psyche sees Amor and falls in love with him. They are separated, and the separation is maintained by Amor’s mother, Venus, thus presenting the standard “familialization” of the pivotal conflict (in which the blocking characters are family members, thus giving their antagonism greater emotional force). Psyche now seeks reunion with Amor. Unlike the Arawak tale, this story develops the quest by the abandoned, human lover. Specifically, Psyche tries to win the approval of Venus by performing various tasks dictated by Venus—a scenario that we might expect for the male lover, rather than the female. (In Chang Boils the Sea, we see a sort of parallel for this, with the male lover undertaking the work of convincing the beloved’s father—in this case antagonistically, rather than deferentially.) In the interim, Amor recovers from an injury caused by Psyche’s misguided attempt to identify and kill him. He manages to procure immortalizing ambrosia. Drinking the ambrosia, Psyche becomes immortal, thus making possible her enduring union with Amor.

Thus, in Apuleius’s story, we have the concealment of identity, the taboo on seeking or revealing that identity, the violation of the taboo, and the separation. This is then followed by the usual familial conflict and a version of the lover’s attempt to convert the recalcitrant parent. However, the resolution does not come from a conversion of the parent. It comes, rather, from a transformation in the status of the lover, a change from mortal to immortal. This change erases the ontological difference that made the union impossible to begin with. There are two points to make in connection with this. First, there are two obvious ways in which the ontological difference may be overcome. One is the elevation of the human lover, as we see here. The other is the degradation of the spirit lover, as in the case of Urvaśī when she is initially joined with Purūravas in the Viṣṇu Purāṇa. The second point is that this is parallel with the more ordinary or basic romantic structure, in which the conflict is social, as when the lovers derive from different classes or castes. The elevation of one lover is, in this respect, parallel with the discovery that, for example, the slave girl is actually a patrician who was kidnapped as a child. The degradation of one lover is parallel to the discovery that the apparently higher class or caste beloved is in fact not inaccessible to the lover—as when Duṣyanta (in Kālidāsa’s Abhijñānaśākuntalam) discovers that Śakuntalā is not from the higher, Brahmin caste, but has only been raised by a Brahmin.

Thus, we find Native American, Chinese, Indian (South Asian), and European instances of the intensified romantic plot in which the lovers violate an ontological constraint defined by an animacy hierarchy. These stories also occur across a range of time periods. Further examples may be found in more recent works, both western and eastern. For example, Wings of Desire (directed by Wim Wenders), presents us with the story of an angel who falls in love with a mortal woman. He loses his angelic status, becoming mortal, and is united with the woman. Here, the sort of concealment we found in the Amor and Psyche story occurs before the union of the lovers, as the angel can observe his human beloved, while the latter has only vague intimations of the angel’s presence. In this case, there is no restriction, restriction-violating revelation of identity, and consequent separation of the lovers; it is, then, a reduced or simplified version of the structure.

Another film with this sort of plot is Postman to Heaven, a 2009 work from South Korea (directed by Hyung-min Lee). In this film, a young woman and a ghost fall in love. The ghost has been assigned the task of helping bereaved men and women work through their grief. The young woman had recently lost a lover, which leads to her connection with the ghost. The twist in this story is that the ghost is visible only to people who are in mourning. Thus, the young woman will no longer be able to see or interact with the ghost from the moment she overcomes her grief. But she overcomes her grief in proportion to the degree to which she falls in love with the ghost. Thus, she loses her new beloved precisely to the degree that she loves him; thus, she suffers a new grief precisely to the degree that she overcomes her old grief. The problem is resolved by divine intervention. Due to the ghost’s success in consoling the bereaved, particularly this young woman, Heaven returns him to his mortal life. As a human, he is now able to be united with his beloved, who is thereby freed from both the old and the new grief.

Neither film has the concealment-taboo-violation structure. But both still alter the more usual, basic genre trajectory, by having the lovers violate an ontological restriction and allowing their union by a change in status, from spirit to human. Further cases present other variations. For example, complex instances may be found in other stories by Pu Songling, such as “Grace and Pine,” “Butterfly,” and “The Laughing Girl.”

Romantic Stories and Same-Sex Love: An Intermediate Case

We will turn to the ethical form of the intensified romantic structure in the next section. Before going on to that, however, we might briefly remark on Ovid’s story of Iphis and Ianthe and the Śikhaṇḍinī story in the Mahābhārata. In both stories—one European, the other Indian—two women are married to each other due to concealed identity, though the concealed identity in this case is one of anatomical sex, not of ontological level (human, animal, or spirit). As with the cases considered thus far, the revelation of the concealed identity causes problems for the union, and in each case the misrepresented lover (Iphis and Śikhaṇḍinī) goes through a magical sex change, comparable to the transformation of a human lover into a spirit or vice-versa.

These cases are interesting in part because they show how the romantic structure may, through parametric variation, be applied to significant social issues, in this case same-sex love. This is not to say that either provides any solution to the social conflicts surrounding such love, any more than the animacy-based versions provide a solution to the love between people of extreme out-groups (as in the Arawak case). The transformation of one lover functions most obviously as a fantasy, perhaps giving some people imaginary satisfaction. In any case, its social function is complex and equivocal. For example, in the case of Iphis and Ianthe, it may be taken to validate lesbian love, indicating that the lovers should be joined. But it might also be taken to stress the necessity of heterosexuality, since they can be joined only when they are differently sexed. On the other hand, romantic stories tend to cultivate a very strong preference for the lovers’ union. Thus, my inclination is to see stories such as Iphis and Ianthe as functioning primarily to cultivate sympathy with the same-sex lovers, even as the ending gives the author and reader an out, enabling the plausible denial of such homoerotic sympathies.

I refer to this as intermediate because it is not clear how authors and readers conceive of the restriction in this case. I suspect that it varies with authors, readers, and cultures. Some would think of the restriction as ethical. I suspect that such readers would not, in general, be sympathetic with the same-sex lovers. Other readers would think of the restriction as simply social, comparable to the forbidding of intermarriage between Montagues and Capulets (this is presumably the view of most academic readers today). Finally, some would think of the restriction as part of the order of being, so to speak—in this case, related not to animacy as such, but to the issue of mating. Different species cannot interbreed; neither can spirits and humans. For some writers and readers, the relation of same-sex lovers is probably of this sort. Clearly, this is a topic that requires further research, with many more examples from a much wider range of traditions.

Romantic Stories and Ethical Restrictions: Incest and Adultery

In considering the ontological version of intensified romance, I could not help but be aware of the similarity to one of the most famous and influential stories in European literature, the story of Oedipus. Though not a romantic plot per se, the story does involve two people being united despite the fact that their union is rigorously forbidden. This is possible because the (familial) identity of one person, Oedipus, has been concealed. Thus no one recognizes that the union is incestuous. When that identity is revealed, the couple is separated. In this case, the results are tragic, with the woman hanging herself and the man blinding himself. Interestingly, as developed by Sophocles (in the fifth century before the common era), the story is not only principally romantic (despite the union of Oedipus and Jocasta), but sacrificial. Specifically, there is a plague in Thebes; it has resulted from Oedipus’s crime of (unknowingly) killing his father, which led to his further crime of (unknowingly) marrying his mother. In other words, in this story, a sin perpetrated by the primary representative of the society (the king) has led to social devastation. The result is the sacrifice of the guilty parties (Jocasta’s death and Oedipus’s blinding). In Oedipus the King, then, a sacrificial structure is integrated with a much more limited romantic structure.

As just noted, the form of identity at issue for Oedipus is familial. Some readers have been confused by this, because the identity issue is not one of humans and animals or spirits. That is right. Indeed, that is the difference between two sorts of intensified romance—the ontological and the ethical. Ontological problems keep people and insects from marrying; ethical restrictions bear rather on incest. Of course, the ontological stories may metaphorically represent moral conflicts—or mere social conflicts. The point here concerns the literal representations.

Ovid’s Metamorphoses (early first century of the common era) presents us with other incestuous romantic narratives. One is the simple story of Byblis, who tries to unite with her brother, but is rejected; weeping, she is transformed into a spring. Another, more interesting story concerns Myrrha and her father. Myrrha falls in love with her father and contrives a way of sleeping with him while concealing her identity. When he learns of her identity, he not only rejects her, but tries to kill her. She flees and is transformed into a tree.  The Byblis story is a simple tale of rejected love. The story of Myrrha, however, includes union with concealed identity, followed by revelation and tragic separation.

One might expect other moral restrictions to enter here as well. An obvious case is adultery. In European literature, works such as Anna Karenina come to mind. In Tolstoy’s novel, for example, we have love between Anna and Vronsky, leading to a partial union, which is concealed from society. Once revealed, inadvertently, by Anna, problems arise. On the other hand, this is not a concealment of identity; Anna does not misrepresent herself as unmarried—either to Vronsky or to the larger world. Rather, it is simply the (illicit) sexual relations that are concealed. Moreover, the revealing of the secret does not lead to the separation of the lovers. There are problems with the cross-cultural manifestation of this structure as well, since possibilities for polygamy or divorce limit the degree to which prior marriage serves as a restriction on the lovers. On the other hand, there are some similarities between the structure we have been examining and such adultery stories as Anna Karenina.

Implications of this Analysis

Clearly, the preceding examples of an ethical romance plot are taken from a single tradition. That may suggest that this specific form of the structure is parochial. However, the Oedipus and Myrrha stories remain cases of a larger sequence that does recur cross-culturally.

Again, the intensified romance is defined by a restriction on the lovers that is not simply a matter of social convention or other social contingency, but of identity bearing on animacy or ethics. In its most prototypical form, the intensified romance proceeds as follows:

• The lovers are united.
• However,  there is some (intentional or unintentional) concealment of the ontologically or ethically relevant identities.
• In connection with this concealment, there is often a warning about the behavior of one or both lovers.
• A violation of the warning results in the problematic identity being revealed.
• As a result of this revelation, the lovers are separated.
• In comic instances of the ontological version, that separation may be overcome when the identity of one lovers is changed.
• However, the ethical version of this structure does not in general allow for such a comic transformation (e.g., Psyche can be made immortal, but Oedipus cannot be made into the son of someone other than Jocasta).

Faced with the preceding examples, we might wonder why there is such a recurring pattern. “Why” here might refer to the generation of the sequence, just what gives rise to it; this could equally be referred to as “how” the structure is produced. “Why” might also refer to the function of the resulting sequence; this is “why” in the sense of “what for.” Note again that the pattern here is a pattern in stories. It is not a pattern in real life (where people do not marry insects either with or without concealed identity). It is, rather, a pattern in simulation.

The “how” question is relatively easy to answer. Again, we may think of simulation as involving principles with parameters and local variables. The simulation of romantic stories involves some restriction that separates the lovers. Obvious possibilities for such separation would include conventional versus necessary, with the latter category including ontological and ethical cases. Within each category, any given character might or might not know about the restrictions or the identities that trigger such restrictions at any given time. (For example, Oedipus and Jocasta do not know their restriction-triggering identities—son and mother—when they get married.) The stories we have been considering begin with an extreme restriction and vary the timing and extent of relevant knowledge. Typically, the knowledge comes to have social consequences only when it is socially available (as when the man’s society learns about the Jaguar woman’s identity). Usually, the lovers are able to maintain their union when they have the relevant knowledge but society does not, though the story of Myrrha suggests that this may be less true in the ethical cases.

The functional significance of this structure is perhaps less clear. However, there are some obvious possibilities. Specifically, we would expect the function to be either emotional or thematic. Emotionally, this sub-genre of romance provides some degree of novelty relative to the basic (and more realistic) form. In connection with this, it often organizes the trajectory of feeling somewhat differently. Specifically, the basic structure commonly begins with the main characters falling in love and recognizing their reciprocal feelings, but then being separated very quickly. The intensified version often involves the lovers joining together in a regularized union, such as marriage, and maintaining that state for some time before being separated due to the discovery of the concealed identity. Both versions treat love, loss, and (in the full, comic version) reunion. But in the basic version, the love is the feverish love of early romance and the loss is a loss of imagined possibilities. In the intensified version, the love is more likely to include a strong “companionate” element, a sense of bonding that may be less intense but may be more clearly enduring, and the loss is a loss not only of imagined possibilities, but of the ordinariness of regular life and ongoing expectation experienced for, in some cases, years of family life.

For the reasons just mentioned, the loss of the beloved in the intensified version may be more similar to the loss of a spouse in death. This too has emotional consequences. Specifically, in its ontological variant, the structure may suggest bereavement, while allegorizing that bereavement in the (perhaps less distressing) form of the human-spirit romance. I have already noted that there are suggestions of this in Vikramōrvaśīyam. The connection seems clear in other works as well. For example, in Postman to Heaven, the young woman falls in love with a ghost. Her former lover has died and her new beloved is actually a boy in a coma in a hospital. The lovers are separated as human and ghost, until his spirit returns to his body and he is revived from the coma, at which point they meet as mortals. The film repeatedly points toward the loss of a long-term partner by death. But it presents the main lovers’ separation as a matter of an ontological incompatibility which is resolved by the spirit leaving heaven to become a mortal.

These points are related to the thematic purposes of the ontological romance. In the comic version, the thematic operation of the structure involves a suggestion that the losses of death may be overcome—a point with obvious emotional consequences as well. In other words, it suggests that apparently irrecuperable losses might in fact be recuperable, and it does this without drawing our attention to death and separation too directly. Indeed, in some cases, it even serves to reconnect our companionate love with earlier, more romantic feelings. I say “our” here because the reader’s relation to the emotions of the protagonists is often intertwined with his or her own emotional memories (see, for example, Oatley, Passionate, 36, 179, 180). In that way, our empathic response to the lovers is in some degree connected with our own romantic histories.

On the other hand, the intensified romantic story is not always related to long-term attachment and bereavement. For example, the Arawak story has suggestions of death and grief. But it also suggests social divisions that have been rendered more extreme by being assimilated to species differences. There are two obvious thematic purposes of this sort of intensification of difference between the lovers. The first is to present an a fortiori case of romantic love in opposition to social divisions. If the Arawak hunter can live an ideal marriage with an actual jaguar or tiger woman, how much more readily could an Arawak live in marital union with another person who is from a class or other society associated with jaguars or tigers? As I have argued elsewhere (see chapter seven of Understanding Nationalism), romantic stories tend to challenge in-group/out-group divisions, preferring the union of the lovers to the maintenance of social oppositions. The ontological romance may simply extend this inclination.

This critical function of the ontological romance is clearest when the lovers’ initial union is untroubled. In the Arawak story, there is no real problem with the lovers’ life together. In other stories, however, there are difficulties, sometimes intrinsic to the lovers’ identities, as in the difference between mortals and immortals. For example, in Wings of Desire, the lovers can be joined only when the difference between them is overcome, when the angel becomes human. This too suggests that ordinary, social divisions are not absolute. If the difference between humans and spirits may be effaced, if immortal spirits can become humans and humans can become immortal spirits, then surely members of one society can become members of another society. This may simply be a matter of one lover being accepted into the other’s society, or it may involve some more elaborate assimilation into the culture of the latter.

Again, the obvious exception here is ethical, as in the case of incest. Incestuous unions are typically not open to this sort of resolution. Oedipus and Jocasta seem perfectly happy together. Everything appears to be going along well, with the family flourishing, including the children. But the union remains wrong for most readers. Thus, the ethical version does not fit with the most obvious thematic purposes of the ontological version of the intensified romantic structure. That may account for its apparent infrequency, relative to the ontological version, both within European tradition and across cultures. On the other hand, the success of the Oedipus story indicates that this version may be emotionally successful.

Before concluding, it is worth noting that concealed identity operates as a motif that may be used in various narrative contexts. Moreover, it may be taken up in complex and surprising ways in relation to romantic narrative. Particularly striking cases of this sort may be found in Shakespeare’s “problem” plays, All’s Well that Ends Well and Measure for Measure. Specifically, in both cases, one main character sets out to engage in adultery, deceptively seducing a young woman, thus initiating a train of events that is commonly developed in the seduction plot (on the seduction plot, see chapter four of my Affective). However, the lovers meet in circumstances where the identity of the woman is concealed (reminiscent of the meeting between Myrrha and her father). However, in these cases, the concealed identity of one of the lovers does not mean that an apparently legitimate union is illegitimate. It means, rather, that an apparently illegitimate union is legitimate, for in both cases the man’s wife takes the place of the woman he was aiming to seduce. The resulting stories are highly interesting, due to this reversal. But they are problematic because the final union is not based on mutual love and it is therefore difficult for the reader (or audience member) to be entirely pleased by the putatively comic conclusion.

In connection with these final examples, one referee of this essay worried that concealed identity may be a motif produced by the romantic genre, not occurring elsewhere. I should therefore mention that in heroic plots we encounter spies who conceal their identities and heirs to the throne who are unknown to be heirs to the throne; in revenge narratives, we find revengers disguising themselves and killing people whose identity they misconstrue; in criminal investigation narratives, detectives take on disguises and innocent people are sometime mistaken for criminals—and so on, for all genres. Perhaps the most relevant case here is the (already mentioned) seduction story, which is often more appropriately termed a rape story. In this structure, the seducer or rapist often conceals his identity. This is significant to the present study because there are some famous cases in which the rapist is a god and the false identity is that of an animal—as in Zeus’s rape of Leda. This is related to the cases we are considering as the union is “impossible,” in the sense we have been considering. It therefore suggests a similar concern with the violation of putatively necessary social divisions, though in another genre.

Future Research

An obvious topic for research in this area concerns ethical restrictions on lovers and the degree to which the topics of incest and adultery, or other possibly morally objectionable unions, do or do not develop a version of this structure. A parallel point applies to same sex unions and the degree to which the relevant restrictions are presented as ethical, ontological, or merely social and conventional. Instances of these relationships and restrictions are clearly under-represented in the preceding discussions.

On the other hand, research on the incidence and varieties of more unequivocal cases of ontological romance is important as well. Again, we find a range of stories in different traditions in which lovers face a restriction that is not merely conventional or contingent, but in some sense viewed as necessary, either ontologically (in relation to animacy) or morally. Prototypically, the lovers in such stories are united, due to some intentional or inadvertent concealment of identity. But that concealment often renders the union unstable. In connection with this, one of the lovers may be warned not to engage in behavior that will reveal the concealed identity. In a tragic error, however, he or she may do just the wrong thing, precipitating the lovers’ separation. In a comic version, one of the lovers’ status may change, allowing the union, but this is generally restricted to the ontological version. We have evidence that the ontological version of this intensified structure occurs in a range of genetically and areally unrelated traditions. But it is not clear how frequent (or infrequent) the appearance of this structure may be. Further research is also important for assessing the preceding conjectures on the generation of the sub-genre and, even more, its function. The generation of the structure—deriving from the operation of variables in principles governing simulation—seems fairly clear in general. However, the specifics of the variables and the resulting variations on the sub-genre could be further developed and clarified. Moreover, the functions of the intensified romance appear more debatable.

Especially in connection with the last point, the relation of this variant to changes in cultural ideology and historical conditions may be worth investigating. For example, one might wonder if the ontological romance occurs more frequently in periods of extensive inter-cultural contact, where the contact has not become routinized; in other words, one might wonder if a particular degree of interaction with alien societies may not promote stories treating the romantic relations of extremely distant partners. For example, we might ask if the Pu Songling stories owe something to the early Qing dynasty experience of Manchu-Chinese cultural conflict (see Roberts 142), the divisive contact with the West and associated anti-Christian reaction of the time (see Rowe 138-139), and related matters. Similarly, we might consider if Apuleius was affected by the cultural interactions of 2^ndCentury Rome, with its extensive empire. Along the same lines, we might inquire into the possible relation between the Arawak tale and European contact. A correlation in this respect might constitute a statistical, implicational universal.

I should note here that I am referring specifically to the ontological version of the intensified romantic structure. Such issues as same-sex desire recur in a range of time periods and cultures presumably because same-sex desire recurs in all time periods and cultures. As one reader of this essay commented, stories of same-sex union (as well as adultery and other sexual behaviors) occur because social rules often relate to categorial identities and these are often in conflict with what I have called “practical identity” (see, for example, Understanding Nationalism 25-37), one’s actual capabilities and propensities. That practical identity may include, for example, same-sex desire. However, the case of ontological romance—and even, for the most part, that of the most extreme cases of ethical romance, such as those involving incest—are different. There is no social issue of people marrying insect spirits or even sons marrying their mothers. These stories are, then, fundamentally a product of simulation. Of course, in the end, our narrative simulations, here and elsewhere, are inseparable from real-world issues of categorial and practical identity, as suggested by the Arawak story’s possible allegorization of social identity conflict—just as our experiences of categorial and practical identity are inseparable from simulation. In this way, the two forms of analysis should ultimately be joined with one another.

Finally, one might wonder about the implications of these observations for other genres. The sacrificial operation of the ethical romance in Oedipus the King might lead us to inquire after the relation of this romantic structure to other genres with which it might be integrated. We might also ask if there are junctural points in those other genres—sacrificial, heroic, family separation and reunion, revenge, seduction, and criminal investigation—that define significant sub-genres, such as we have been considering in the case of romantic stories. More exactly, simulation is not confined to romance, but produces the other genres equally. As such, we would expect to find such complex, sub-generic patterns in those other genres as well. In some cases, such sub-genres will be related to the intensified romance, with its ontological and moral impossibilities. Stories such as the rape of Leda suggest that this may be found in seduction/rape structures. Alternatively, such a violation may serve as a recurring motif. For example, there is an element of this in the story of Mary’s conception of Jesus—part of a sacrificial narrative—with its divine/human union and its image of the paraclete as a dove.

[See also Nigel Fabb, “Comments on Hogan, ‘Impossible Love and Henry Fielding’s Tom Jones’” and Geoffrey Russom, “Comments on Hogan, ‘Impossible Love’.”]

Works Cited

Aikhenvald, Alexandra. A Grammar of Tariana. Cambridge: Cambridge University Press, 2006.

Hogan, Patrick Colm. Affective Narratology: The Emotional Structure of Stories. Lincoln, NE: University of Nebraska Press, 2011.

Hogan, Patrick Colm. How Authors’ Minds Make Stories. Cambridge: Cambridge University Press, 2013.

Hogan, Patrick Colm. The Mind and Its Stories: Narrative Universals and Human Emotion. Cambridge: Cambridge University Press, 2003.

Hogan, Patrick Colm. Understanding Nationalism: On Narrative, Identity, and Cognitive Science. Columbus, OH: Ohio State University Press, 2009.

Kālidāsa. Theater of Memory: The Plays of Kdlidasa. Ed. Barbara Stoler Miller. Trans. Edwin Gerow, David Gitomer, and Barbara Stoler Miller. New York: Columbia University Press, 1984.

Lee, Hyung-min, dir. Postman to Heaven. Written by Eriko Kitagawa. Seoul, South Korea: CJ Entertainment, 2009.

“Legend of the White Snake.” Shanghai Daily(26 March 2017). Available at https://www.shine.cn/archive/sunday/now-and-then/Legend-of-the-White-Snake/shdaily.shtml (accessed 21 June 2018).

Lévi-Strauss, Claude. From Honey to Ashes (Introduction to a Science of Mythology: 2). Trans. John and Doreen Weightman. New York: Harper & Row, 1973.

Li Hao-ku. Chang Boils the Sea. In Six Yüan Plays. Trans. Liu Jung-en. New York: Penguin, 1972, 159-187.

Oatley, Keith. “Meetings of Minds: Dialogue, Sympathy, and Identification, in Reading Fiction.” Poetics26.5 (1999): 439-454.

Oatley, Keith. The Passionate Muse: Exploring Emotion in Stories. Oxford: Oxford University Press, 2012.

Pu Songling. Strange Tales from a Chinese Studio. Trans. John Minford. New York: Penguin, 2006.

Roberts, J. A. G. A Concise History of China. Cambridge, MA: Harvard University Press, 1999.

Roth, Walter. An Inquiry into the Animism and Folk-Lore of the Guiana Indians. Thirtieth Annual Report of the Bureau of American Ethnology to the Secretary of the Smithsonian Institution: 1908-1909. Washington: Government Printing Office, 1915, 103-453.

Rowe, William. China’s Last Empire: The Great Qing. Cambridge, MA: Harvard UP, 2009.

Wenders, Wim, dir. Wings of Desire. Written by Wim Wenders, Peter Handke, and Richard Reitinger. Berlin, West Germany: Road Movies Filmproduktion, 1987.

Wilson, H. H. “Introduction.” In Vikrama and Urvashi, or The Hero and the Nymph. Trans. H. H. Wilson. Calcutta, India: Society for the Resuscitation of Indian Literature, 1901, i-xii.

Zoltán Kövecses, Eötvös Loránd University, Hungary

In the paper, I will be concerned with the issue of metaphor universals in poetry only – leaving aside the genre of the novel, for example (but see Hogan, The Mind). This decision is the result of practical considerations, rather than any theoretical reasons. Within the framework of conceptual metaphor theory, a lot more work has been done on poetry than on the novel, probably mostly because a pioneering study on metaphor by Lakoff and Turner (More than Cool Reason) also focused on poetry.

By metaphor I mean conceptual metaphor, which consists of a set of systematic mappings between two domains of experience (Lakoff and Johnson, Metaphors We Live By). One domain, which is typically more abstract, is called the target domain, and the other, which is typically more concrete or physical, is called the source domain. The more concrete domain is used to understand the more abstract one. A conceptual metaphor can have a variety of different manifestations in various modalities, including, most importantly for the present purposes, linguistic manifestations. In other words, a distinction is made between metaphor as a conceptual pattern (given structure by the mappings) and the linguistic manifestations (or examples) of this conceptual pattern (for overviews, see, e.g., Lakoff, “The Contemporary Theory”; Kövecses, Metaphor).

Metaphor Universals and Universal Metaphors

We should make a distinction between metaphor universals and universal metaphors. The notion of metaphor universals is more general than that of universal metaphors. Metaphor universals can be conceived of as any metaphor-related phenomenon, including universal metaphors. For example, one metaphor universal is the use of metaphors by poets. Probably, there is no poetry without the use of metaphors in some form. It is an interesting question why this should be the case. The answer favored by cognitive linguists would most probably be that typical topics of poetry, such as love, freedom, beauty, history, time, life, honor, nature, suffering, and so on, all invite metaphoric conceptualization, as they are highly abstract concepts that make excellent target domains in conceptual metaphors.

Universal metaphors, on the other hand, are the conceptual metaphors that are used universally, near-universally or potentially universally. Conceptual metaphors such as love is fire, time is motion, or life is a journey are potential universal conceptual metaphors in everyday linguistic usage, but also in literature, as poetry in many unrelated languages around the world shows.

The Cognitive-Linguistic Study of Conceptual Metaphors in Poetry

The study of conceptual metaphors in cognitive linguistics began with Lakoff and Turner’s book, More than Cool Reason. Lakoff and Turner made two very important claims concerning conceptual metaphors in poetry. First, they showed that poets share with everyday people most of the conceptual metaphors they use in poetry. The reason for this is that the conceptual metaphors such as the ones mentioned above are based on shared bodily experiences – for non-poets and poets alike. (More about this later.) Second, Lakoff and Turner suggested that metaphorical creativity in poetry is the result of four common conceptual devices that poets use in manipulating otherwise shared conceptual metaphors. These include the devices of elaboration, extension, questioning, and combining. In other words, according to Lakoff and Turner, the conceptual metaphors will be (nearly) the same, but the linguistic manifestations reflecting the effects of these devices will (or can) be (somewhat) different in everyday and poetic forms of language.

However, others have shown that these four cognitive devices, or strategies, exist not only in poetic language but also in more ordinary forms of language use, such as journalism (see, e.g., Jackendoff and Aaron, Review Article; Semino, Metaphor in Discourse). Thus, on this basis alone, it is not possible to distinguish poetic from non-poetic metaphor. Moreover, in accounting for poetic metaphors, Turner proposed that in many cases poetry (and literature in general) makes use of what he and Fauconnier call “blends,” in which various elements from two or more domains, or frames, can be conceptually fused, or integrated (see, e.g., Turner, The Literary Mind; Fauconnier and Turner, The Way We Think).

Although many conceptual metaphors are shared by poets and non-poets, many are clearly not. These are metaphors that are not based on universal bodily experiences but on certain creative analogies between a source and a target domain set up either by lay people or poets. In these cases, we have certain resemblances between a source and a target. While the dominant trend in the cognitive linguistic study of metaphor is to deal with metaphors where the source and target are related by some correlation in experience, analogical relations between a source and a target build on similarities, or resemblances, of various sorts (real, physical similarities, generic-level similarities, imagined similarities, etc.). They range from simple to complex. Below is a simple one that comes from William Wordsworth’s poem “I Wandered Lonely as a Cloud”:

I wandered lonely as a cloud
That floats on high o’er vales and hills…

A more complex example is from Shakespeare’s King John. The king says the following to a messenger whose face reveals that he has some bad news to tell him:

So foul a sky clears not without a storm.
Pour down thy weather.

We can spell out the similarities in this example as a set of mappings between the scene of an imminent rainstorm and the scene of a messenger just about to deliver a message to the king:

the appearance of the sky ⇒ the appearance of the messenger’s face
the imminent storm ⇒ the bad message likely to be delivered
the rain ⇒ the act of telling the bad news

We can take similarity-based metaphors (i.e., analogies or similes) as a special case of conceptual metaphors (besides correlation-based ones). Needless to say, such potential similarities can be used to create a huge number of cases in metaphorical expression that are anything but universal.

Metaphorical Universality and Non-Universality in Poetry

Given the rough sketch of conceptual metaphors above, it seems that certain poetic metaphors (those that are correlation-based) are (potentially) universal at the conceptual level but non-universal at the linguistic one. According to the “standard” account of conceptual metaphors (see, e.g., Lakoff and Johnson, Metaphors We Live By; Lakoff, “The Contemporary Theory”), conceptual metaphors are based on fundamental bodily experiences (i.e., correlations in experience). It is this basic bodily experience that may potentially lead to universal conceptual metaphors in poetry.

A more refined version of this idea in cognitive linguistics can be found in the theory of “primary metaphors” (see Grady; Lakoff and Johnson, Philosophy in the Flesh). On this view, there are correlations between certain sensorimotor and subjective, abstract experiences, such as between destinations and purposes, body heat and emotion, verticality and amount, and so on. The repeated experience of such correlations results in primary metaphors that are, at least potentially, universal, including purposes are destinations, intimacy is closeness, emotion is heat, more is up. The primary metaphors constitute the basis for complex or compound conceptual metaphors such as life is a journey and love is a unity. For example, purposes are destinations is one primary metaphor that, in part, conceptually constitutes life is a journey, intimacy is closeness is one that, in part, constitutes love is a unity, and emotion is heat is one that, in part, constitutes anger is a hot fluid in a container. As a result, these pairs of primary and compound metaphors are potentially universal. Indeed, the anger is a hot fluid and the anger is heat conceptual metaphors are found in dozens of genetically unrelated and spatially and culturally independent languages, such as English, Chinese, and Hungarian (see, e.g., Yu, A Perspective from Chinese; Kövecses, Metaphor and Emotion, Metaphor in Culture, Metaphor). It would not be surprising to find linguistic metaphors based on these and other similar body-based conceptual metaphors in the poetry of these and other languages. As an example, let us take a poem by Anne Bradstreet, a 17^th-century American poet, who in her poem “To My Dear and Loving Husband” wrote:

If ever two were one, then surely we.
If man were loved by wife, then thee

The first line is a straightforward linguistic illustration of the conceptual metaphor love is a unity (of two complementary parts) (see Kövecses, The Language of Love). It is very likely that the unity metaphor for love can be found in the poetry of many languages and cultures around the world.

While the view briefly described above provides an elegant and coherent account of universality in everyday and poetic metaphors, it does not pay sufficient attention to the many cases of non-universality. Non-universality in metaphor is complementary to universality; one assumes the other. Since we already provided an answer to the question of what makes some conceptual metaphors universal, we can take this as our point of departure in attempting to answer the next question: What makes some other conceptual metaphors non-universal? Or putting the question differently: Where does non-universality in everyday and poetic metaphorical thought come from?

In recent years, I suggested that it is necessary to make certain adjustments to the standard theory of conceptual metaphors in order to be able to answer this question (see Kövecses, Metaphor in Culture, for some initial ideas). The adjustments concern the role of context in the creation of novel metaphors (both conceptual and linguistic), on the one hand (see Kövecses, Where Metaphors Come From), and the various degrees of schematicity of conceptual metaphors (and their linguistic expressions), on the other (see Kövecses, “Levels of Metaphor”). We can term the resulting conception a contextualist and multi-level view of conceptual metaphors.

Both adjustments concern, in part, the issue of why conceptual and linguistic metaphors diverge from the universal patterns suggested above. They point to two major forms of deviation from universality. First, the insistence on the role of context helps us account for the kind of metaphor variation that derives from individual and group differences in human experience. And second, the multi-level view can explain the divergences in the use of metaphor resulting from individual and group differences in how higher-level conceptual metaphors are elaborated by lower-level ones. Actually, the two processes are related: various contextual factors can trigger particular elaborations of higher-level conceptual metaphors. This interaction will be demonstrated below.

Since deviations from universality occur on a large scale, we should not consider the issue of universality to be the main focus of metaphor research, as is often the case by cognitive linguists. As was noted above, universality and variation complement and assume each other. In conceptual metaphor theory, we should pay just as much attention to variation as to universality (see Kövecses, Metaphor in Culture), since variation is pervasive in the use of metaphor. This, I suggest, is due, to a large extent, to the role that context plays in metaphorical conceptualization – both in everyday usage and poetry.

How Context Shapes Metaphorical Conceptualization

Based on my findings (Kövecses, Where Metaphors Come From), I suggest that there are a variety of contextual factors (to be discussed below) that prime speakers when they use metaphors in communicative situations. The contextual factors belong to several context types: situational context, discourse context, conceptual-cognitive context, and bodily context. These types of context each come in two forms: local context and global context, as shown in Figure 1 below. The local context involves the specific knowledge conceptualizers have about some aspect of the immediate communicative situation. By contrast, the global context consists of the conceptualizers’ general knowledge concerning their community. It involves knowledge shared by an entire community of speakers / conceptualizers. The distinction is mostly of theoretical nature. In many actual communicative situations, there is no sharp dividing line between the local and the global context. The Figure below presents the four major context types as four sections of a circle and their division into local context (as the inner circle) and global context (as the outer circle). The box in the middle represents a particular act of metaphorical conceptualization in context. The contextual factors subsumed under the context types are not given in the Figure. They are discussed in the text following Figure 1.

Figure 1: Summary of Types of Context (Concentric circles represent global and local context; quadrants represent situational, discourse, bodily, and conceptual-cognitive contexts for a given metaphor.)

The four types of context can be briefly described in the following way.

Situational Context

The situational context comprises a variety of different contextual factors. Most commonly this type of context can be thought of as including the physical environment and the social-cultural situation. The physical environment includes the flora, the fauna, the landscape, the temperature, the weather, perceptual properties of the situation, and so on. For example, it is a common observation that American English metaphors relating to the physical environment are characteristically different from those of other English-speaking countries (see, e.g., Kövecses, American English). The small-scale, local environment, such as the visible events in or the perceptual properties of a situation, can also make its influence felt in shaping metaphors. The social-cultural situation consists of social aspects of life that typically center around notions such as gender, class, politeness, work, education, social organizations, social structure, art and entertainment, and others. All of these can play a role in metaphorical conceptualization. For example, Kolodny (The Lay of the Land, The Land Before Her) shows that American men and women developed very different metaphorical images for what they conceived of as the frontier in America. While the women commonly thought of the American frontier as a “garden to be cultivated,” men conceptualized it as “virgin land to be taken.” (For several other examples, see Kövecses, Metaphor in Culture.)

As an example of how the immediate physical situation as part of the physical context can prompt, or prime, a poet to use a particular conceptual metaphor, consider the poem “Dover Beach” by Matthew Arnold. In the first stanza, we get a glimpse of what the physical context consists of:

The sea is calm to-night.
The tide is full, the moon lies fair
Upon the straits,- on the French coast, the light
Gleams and is gone; the cliffs of England stand,
Glimmering and vast, out in the tranquil bay.
Come to the window, sweet is the night-air!

The speaker (poet) is looking out to the sea from inside a house and can see a variety of things. One of them is that “The tide is full.” In the last stanza, however, he can see the tide “retreating” with a “withdrawing roar,” exposing the bottom of the sea:

The sea of Faith
Was once, too, at the full, and round earth’s shore
Lay like the folds of a bright girdle furl’d.
But now I only hear
Its melancholy, long, withdrawing roar,
Retreating, to the breath
Of the night-wind, down the vast edges drear
And naked shingles of the world.

At this point Arnold introduces the conceptual metaphor (christian) faith is the sea (“The sea of Faith”) in a special way. As the sea retreats, it leaves the bottom of sea exposed, and, in the same way, human beings become unprotected by Christian faith that once protected them. It is fairly safe to assume that what induced, or primed, the poet to use this conceptual metaphor in this particular way was the sight before him: the physical event of the ebb and flow of the sea, that is, a contextual factor belonging to what was called the situational context.

Discourse Context

The discourse context involves the immediate linguistic context (i.e., cotext), the previous discourses on the same topic, and the dominant forms of discourse related to a particular subject matter. We’ll look at an example for the linguistic context below. The metaphors in one discourse can also derive from previous discourses on the same topic. This can take a variety of forms ranging from elaborating, extending, questioning, negating, reflecting on, ridiculing, to otherwise taking advantage of a metaphor previously introduced. For example, an MP in the British Parliament responded to the then Prime Minister Tony Blair who said he does not have a reverse gear (i.e., he can only go forward – basing the statement on the progress is motion forward metaphor) with the reply: “but when you’re on the edge of a cliff it is good to have a reverse gear” (example taken from Semino). This was a humorous twist induced by the prior discourse on the progress is motion forward conceptual metaphor. The dominant forms of discourse and intertextuality can also influence the production of metaphors. Since certain forms of discourse can acquire dominant status in a community, the metaphors used in or based on this discourse can become widespread both temporally (historically) and spatially (cross-culturally). For example, the discourse of Christianity commonly gives rise to the use of metaphors in the Christian world.

To see how the immediate linguistic context can produce metaphors in poetic discourse, let us turn to Sylvia Plath’s poem, “Medusa.” Here are some relevant lines:

Off that landspit of stony mouth-plugs,
Eyes rolled by white sticks,
Ears cupping the sea’s incoherences,
You house your unnerving head—God-ball,
Lens of mercies,

Your stooges
Plying their wild cells in my keel’s shadow,
Pushing by like hearts,
Red stigmata at the very center,
Riding the rip tide to the nearest point of
departure,

Dragging their Jesus hair.
Did I escape, I wonder?

In the world of the poem, Sylvia Plath addresses her mother as Medusa. In Greek mythology, Medusa is a gorgon with snakes for her hair, who turns people who look at her to stone. As the lines quoted above suggest, the poet is trying to escape from the harmful influence of the mother. (This can be seen most clearly in the line “Did I escape, I wonder?”). What is remarkable here is that, to convey this, the poet makes use of the other sense of medusa: the “jellyfish” sense (“Your stooges / Plying their wild cells in my keel’s shadow”). She’s trying to get away from an overbearing mother, and the mother is portrayed analogically as jellyfish. Schools of jellyfish move about in the sea, and jellyfish stings can inflict pain and even death in humans. Thus, it can be suggested that the “jellyfish” meaning of medusa is used by the poet because the mythological Medusa was introduced early on in the poem (in the title) to begin with. This is one kind of linguistic context – the close relationship between two senses of a word, that is, their polysemy. The word form medusa evokes all the knowledge structures associated with it (given as the two senses of the word), and the poet is taking advantage of them, as they analogically fit the nature of the relationship with her mother.

Bodily Context

A particular state or condition of the body can produce particular metaphorical conceptualizations in specific cases, such as a poet’s or writer’s illness. I call this the bodily context. People’s bodily specificities can influence which metaphors they use. Casasanto (“Embodiment of abstract concepts”) found that left-handers prefer to use the moral is left, as opposed to the moral is right conceptual metaphor. Such metaphors contrast with the metaphors that evolve on the basis of the general properties of the human body (i.e., the correlation-based primary metaphors). Given the capacity of body specificities to prime metaphorical conceptualization, we can take the body as a further type of context[1]. Thus, the body is not only responsible for the production of hundreds of conceptual metaphors through the many correlations in subjective and sensorimotor experience (cf. Grady; Lakoff and Johnson, Philosophy in the Flesh), but it can also prime the use of particular metaphors in more immediate, local contexts (see, e.g, Gibbs; Gibbs and Colston; Boroditsky; Boroditsky and Ramscar).

The idea that the general physical, biological, mental, emotional, etc. condition, or situation, of a poet can influence the way a poet writes poetry is well known and is often taken into account in the appreciation of poetry. Dickinson is a well-studied case, as discussed by several authors, among them, James Guthrie. Guthrie has this to say on the issue:

… I propose to concentrate on the fact of illness itself as a governing factor in Dickinson’s development as a poet. We are already accustomed to thinking about ways in which illness or deformity modulate the registers of expression we hear while reading Milton, Keats, Emily Bronte, Lord Byron. For Dickinson, illness was a formative experience as well, one which shaped her entire poetic methodology from perception to inscription and which very likely shook the foundations of her faith. Reading Dickinson’s poems in the full knowledge and belief that, while writing them, she was suffering acutely from a seemingly irremediable illness renders many of them recuperable as almost diaristic records of a rather ordinary person’s courageous struggle against profound adversity. (Guthrie, Emily Dickinson: 4-5)

Elsewhere, I showed how Dickinson’s (unconscious) choice of metaphors may have been influenced by her optical illness (see Kövecses, “A New View”, Where Metaphors Come From).

Conceptual-Cognitive Context

The conceptual-cognitive context includes the metaphorical conceptual system at large, knowledge about the elements of discourse, ideology, knowledge about past events, interests and concerns. In the metaphorical conceptual system, many concepts (sources and targets) stand in a metaphorical relationship with one another (e.g., life is a journey, argument is war) in long-term memory. Given such metaphorical relationships between concepts (such as between, say, life and journey), their presence or absence in the metaphorical conceptual system may function as a precondition for the production and comprehension of particular metaphors. A metaphorical conceptual system can function as context in this sense. Furthermore, conceptualizers often rely on their knowledge concerning the main elements of a discourse: the speaker, hearer, and the topic. Ideology can also be a formative factor in how metaphors are used in discourse. One’s ideology concerning major social and political issues may govern the choice of metaphors (as work by, for instance, Goatly, Washing the Brain, shows). A good example of this is George Lakoff’s (Moral Politics) study of American politics, where conservatives tend to use the nation is a strict father family metaphor, while liberals prefer the nation is a nurturant parent family version of the generic metaphor the nation is a family. Knowledge about past events (i.e., items in short-term and long-term memory) shared by the conceptualizers may also lead to the emergence of specific metaphors in discourse. For instance, it has often been observed that the memory of historical events can lead to the production (and comprehension) of some metaphors (see, e.g., Deignan, “Metaphorical expressions”; Kövecses, Metaphor in Culture).

Consider the life is a journey metaphor. It can be found in the poetry of many authors, epochs, and traditions. One of these many outstanding poets is Dante, who uses it in his “Divine Comedy”:

Nel mezzo del cammin di nostra vita
mi ritrovai per una selva oscura,

(In the middle of life’s road
I found myself in a dark wood,)

Another is Robert Frost in his poem “The Road Not Taken”:

Two roads diverged in a wood, and I—
I took the one less traveled by,
And that has made all the difference.

The two uses of the life is a journey metaphor are of course not the same. Employing Lakoff and Turner’s terminology, we can say that although the two poets share the same conceptual metaphor, the metaphor is extended differently by Dante and Frost. And the extensions result in different manifestations of the conceptual metaphor: loss of goals in life by the former and making choices in life by the latter.

Elaborations of Higher Level Conceptual Metaphors

Lakoff and Turner (More than Cool) had in mind the kind of elaboration in the previous two examples where the concept of journey is realized lexically in two ways: in one, there is journey through a dark wood and one, where the road of the journey includes a fork that requires a choice. But there is another way of thinking about elaboration. In it, conceptual metaphors may be related to one another at a variety of different levels of schematicity (see Kövecses, “Levels of metaphor”). I suggested that we need to distinguish at least four levels of schematicity in conceptual metaphors: the level of image schemas, that of domains, that of frames, and that of mental spaces.

An example is provided by the American poet Karl Sandburg. Consider the first stanza of the poem called “Skyscraper”:

BY day the skyscraper looms in the smoke and sun and
has a soul.
Prairie and valley, streets of the city, pour people into
it and they mingle among its twenty floors and are
poured out again back to the streets, prairies and
valleys.
It is the men and women, boys and girls so poured in and
out all day that give the building a soul of dreams
and thoughts and memories.
(Dumped in the sea or fixed in a desert, who would care
for the building or speak its name or ask a policeman
the way to it?)

The metaphoric-metonymic use of the skyscraper is clearly prompted by the situational (more precisely, physical-cultural) context. The poem was written in 1916 in Chicago, and it was at the turn of the 20^th century that skyscrapers began to be built on a large scale in major American cities, including Chicago.

But what is more important in the present connection is that the conceptual metaphor on which the image of the skyscraper is based is that of society is a building and that this conceptual metaphor is part of a schematicity hierarchy. At the level of image schemas, we conceptualize complex abstract systems (such as societies) as complex physical objects. At the domain level, societies are conceptualized as buildings and the creation of a society as the physical creation of a building. The latter is the main focus of the metaphor in the poem. At the frame level, the idea of creation is elaborated as construction, which results in the conceptualization of the construction of a society as the construction (i.e., building) of a building (with tools and ingredients). Finally, at the mental spaces level, the building of a new American society is conceptualized as building a skyscraper (with hammers and crowbars and spikes and girders).

Clearly, this is a progression from the most schematic to the least schematic (or most specific) level of conceptualization (or, the other way around, from the most specific to the most schematic level). Thus, we get a set of conceptual metaphors in a schematicity hierarchy of conceptual metaphors:

Level of Image schema:
Complex abstract systems are complex physical objects

Level of Domain:
Society is a building; the creation of a society is the physical creation of a building

Level of Frame:
The construction of a society is the building of a building (with tools and ingredients)

Level of Mental spaces:
Building a new American society is building a skyscraper (with hammers and crowbars and spikes and girders)

This is a schematicity hierarchy that, moving downward, ends in the concept of building a skyscraper. The two highest levels, those of image schemas and domains, may be universal (or at least widespread) in cultures when it comes to the metaphorical conceptualization of societies. But the frame level, and especially the lowest mental spaces level are fairly culture- and even individual-specific. Sandburg captures the idea of building a new American society through the idea of building a skyscraper. At this level of conceptual metaphor, other poets may diverge widely (though using the same conceptual metaphors at the higher levels), depending on a large number of contextual factors (such as the physical environment). In other words, we can conclude on the basis of this example that elaborations of high-level conceptual metaphors may contribute to metaphor variation in poetry as well. Metaphors that may be universal or widespread at higher levels of schematicity may turn into something that is specific to particular poetic traditions or to individual poets.

Conclusions

Are there any metaphor universals in literature? I could not answer this question in the paper because I focused attention on poetry alone. As regards poetry, we can conclude from the foregoing that metaphor is a near-universal feature of poetry (in that most poetry operates with metaphor), or, as Hogan puts it in “What Are Literary Universals?,” a “statistical universal.”

A second metaphor universal in poetry (but possibly also in much of literature in general) is that the metaphors are conceptual metaphors expressed by certain linguistic manifestations. The conceptual metaphors emerge either from universal bodily experiences (correlation metaphors) or from resemblances of various sorts (analogies).

Third, the (unconscious) selection of the metaphors used by poets can be influenced by the various types of context: situational, discourse, bodily, and conceptual-cognitive.

Fourth, the metaphors occur on various levels of schematicity. Four such levels have been distinguished: the level of image schemas, domains, frames, and mental spaces. While this may be a universal feature of poetry, the actual elaboration of higher level metaphors at the most specific level of schematicity (i.e., mental spaces) may produce metaphor variation.

Fifth, the conceptual metaphors that are based on universal correlations in experience are potentially universal; they are present in the poetry of certain unrelated languages/cultures, but not in that of all languages/cultures (i.e., they are not absolute universals). At the same time, the conceptual metaphors based on resemblance tend to be highly variable cross-linguistically.

The issue of metaphor universals goes hand in hand with the issue of metaphor variation. The former often assumes the latter and the latter often assumes the former; they complement each other.

First, metaphor variation is produced essentially by three forces: divergences in the resemblances observed, contextual influence, and different elaborations of higher level metaphors.

Second, even universal metaphors can vary from culture to culture or individual to individual. This is because the universal metaphors may undergo differential contextual influence and because the universal metaphors may be elaborated in different ways. How they are elaborated may also be the result of contextual influence.

Third, metaphor universality is often a matter of the level of schematicity. Higher levels tend to be more universal than lower ones.

Fourth, differences in the kinds of analogies drawn (i.e., divergences in similarities observed) is probably the most obvious and most common source and form of variation.

Finally, it appears that the discussion of metaphor universals and variation in the paper does not allow us to distinguish the use of metaphor in poetry from its use in everyday language. All the metaphor-related processes mentioned above apply to both poetry and everyday language. It seems that the basis for any distinction should be sought elsewhere. It may well be that the proper basis for such a distinction can only be the degree of creativity (novelty), complexity, and density of metaphor (Kövecses, “A New View”). Metaphorical universality and variation are general properties of the human mind, regardless of the domain (everyday vs. poetic) in which metaphors are used.

Future Research

As regards future work in the study of metaphor universals in literature, several things can and should be done. First, researchers should propose accounts of metaphor universality and variation in other genres of literature, including the novel. We should find out whether the view of metaphorical universality and variation as presented in this paper is a viable option for the other genres. Second, a huge amount of work awaits us to collect, analyze, and compare data concerning metaphor universality and variation in diverse languages and cultures around the world. We can propose reasonable hypotheses in this regard on the basis of the CMT approach. Third, and probably most relevantly to literary scholars, we need to figure out whether the findings about metaphor universality and variation as discussed here can contribute useful ideas to literary scholarship in the interpretation of poetry (and other genres). My hope is that the study of metaphor universality and variation along the lines suggested in the paper can enrich accounts of interpretation. At the very least, I hope that the approach I described is not incompatible with certain ways of accounting for poetic meaning, as the comment on the present paper seems to indicate.

[See also Kathleen Hart, “Comment on ‘Zoltán Kövecses, ‘Metaphor Universals in Literature.'”]

Works Cited

Boroditsky, Lera. “Does Language Shape Thought? Mandarin and English Speakers’ Conception of Time.” Cognitive Psychology 43 (2001): 1−22.

Boroditsky, Lera and Michael Ramscar. “The Roles of Body and Mind in Abstract Thought.” Psychological Science 13.2 (2002): 185-189.

Casasanto, Daniel. “Embodiment of Abstract Concepts: Good and Bad in Right and Left Handers.” Journal of Experimental Psychology: General 138.3 (2009): 351–367.

Deignan, Alice. “Metaphorical expressions and culture: An indirect link.” Metaphor and Symbol 18.4 (2003): 255–271.

Fauconnier, Gilles and Mark Turner. The Way We Think. New York: Basic Books, 2002.

Gibbs, Raymond W. Embodiment and Cognitive Science. New York: Cambridge University Press, 2006.

Gibbs, Raymond W. and Herbert Colston. Interpreting Figurative Meaning. New York: Cambridge University Press, 2012.

Goatly, Andrew. Washing the Brain. Amsterdam: John Benjamins, 2007.

Grady, Joseph. “Theories are buildings Revisited.” Cognitive Linguistics 8 (1997): 267-290.

Guthrie, James R. Emily Dickinson’s Vision. Illness and Identity in her Poetry. Gainesville, FL: The University Press of Florida, 1998.

Hogan, Patrick Colm. The Mind and its Stories. Cambridge: Cambridge University Press, 2003.

Jackendoff, Ray and David Aaron. Review Article: More than Cool Reason: A Field Guide to Poetic Metaphor by George Lakoff and Mark Johnson. Language 67.2 (1991): 320–328.

Kolodny, Annette. The Land Before Her. Fantasy and Experience of the American Frontiers, 1630-1860. Chapel Hill: The University of North Carolina Press, 1984.

Kolodny, Annette. The Lay of the Land. Metaphor as Experience and History in American Life and Letters. Chapel Hill: The University of North Carolina Press, 1975.

Kövecses, Zoltán. American English. An introduction. Petersburg, Canada: Broadview Press, 2000.

Kövecses, Zoltán. The Language of Love. Lewisburg: Bucknell University Press, 1988.

Kövecses, Zoltán. “Levels of Metaphor.” Cognitive Linguistics 28.2 (2017): 321-347.

Kövecses, Zoltán. Metaphor. A Practical Introduction. Oxford: Oxford  University Press, 2002/2010.

Kövecses, Zoltán. Metaphor in Culture. Universality and Variation. Cambridge: Cambridge University Press, 2005.

Kövecses, Zoltán. “A New Look at Metaphorical Creativity in Cognitive Linguistics.” Cognitive Linguistics, 2010.

Kövecses, Zoltán. Where Metaphors Come From. Reconsidering Context in Metaphor. New York: Oxford University Press, 2015.

Lakoff, George. “The Contemporary Theory of Metaphor.” In Metaphor and Thought. Ed. A. Ortony. Cambridge: Cambridge University Press, 1993, 202-251.

Lakoff, George. Moral Politics. Chicago: The University of Chicago Press, 1996.

Lakoff, George and Mark Johnson. Metaphors We Live By. Chicago: The University of Chicago Press, 1980.

Lakoff, George and Mark Johnson. Philosophy in the Flesh. New York: Basic Books, 1999.

Lakoff, George and Mark Turner. More Than Cool Reason. Chicago: The University of Chicago Press, 1989.

Semino, Elena. Metaphor in Discourse. Cambridge: Cambridge University Press, 2008.

Turner, Mark. The Literary Mind. New York: Oxford University Press, 1996.

[1] At a meta-level of analysis, it can of course be suggested that assigning different values to left and right is a universal. But the analysis I conduct here is not at the meta-level. At the same time, I recognize the possibility of legitimately identifying universals in this particular case, as well as in other cases mentioned in the paper.