2.1Mental images

It may be recalled that a complete account of verbal metaphor takes on board not only conventional or familiar metaphors, such as the one in (1), but also novel, extended, creative metaphors of the sort in (2). Relevance-theoretic inspiration for the discussion of creative metaphors is credited to Carston (2010Carston, Robyn 2010 “Metaphor: Ad Hoc Concepts, Literal Meaning and Mental Images.” Proceedings of the Aristotelian Society 110: 295–321. , 2018Carston, Robyn 2018 “Figurative Language, Mental Imagery and Pragmatics.” Metaphor and Symbol 33: 198–217. ). The two types of metaphor do not receive a unitary treatment; whereas metaphorical interpretation is ostensively communicated in both cases, and, in this sense, involves recovering the speaker’s intention in context (based on propositions literally expressed), this recovery is not achieved via the same inferential route.

In this connection, understanding the explicit content of the conventional metaphor in (1) requires a quick, local inferential process of adjusting the strict literal sense of a word, i.e. chameleon, to the metaphorical sense of a person with an adaptable disposition, i.e. CHAMELEON*;33.In alignment with the relevance-theoretic tradition, words in capital letters with an asterisk (*) are used to mark ad hoc concepts, whereas words in capital letters without an asterisk represent word meanings outside contexts. a process discussed in terms of ad hoc concept construction (Carston 2002Carston, Robyn 2002 Thoughts and Utterances. Oxford: Blackwell. , 2004Carston, Robyn 2004 “Explicature and Semantics.” In Semantics: A Reader, ed. by Steven Davis, and Brad Gillon, 1–44. Oxford: Oxford University Press., 2013Carston, Robyn 2013 “Word Meaning, What Is Said and Explicature.” In What Is Said and What Is Not, ed. by Carlo Penco, and Fillipo Domaneschi, 175–204. Stanford: CSLI Publications.; Wilson and Carston 2007Wilson, Deirdre, and Robyn Carston 2007 “A Unitary Approach to Lexical Pragmatics: Relevance, Inference and Ad Hoc Concepts.” In Pragmatics, ed. by Noel Burton-Roberts, 230–259. Basingstoke: Palgrave. ). On the other hand, a distinct type of inferential route is required for the derivation of the metaphorical content of a creative metaphor, such as (2). This type of route typically implicates a slower, global appraisal or metarepresentation of the literal content of the metaphor (the metaphorically intended part of the text) as a whole.

Carston (2010Carston, Robyn 2010 “Metaphor: Ad Hoc Concepts, Literal Meaning and Mental Images.” Proceedings of the Aristotelian Society 110: 295–321. , 2018Carston, Robyn 2018 “Figurative Language, Mental Imagery and Pragmatics.” Metaphor and Symbol 33: 198–217. ) explains the nature of creative metaphor interpretation by way of illustration. One of the examples discussed is the following passage from a novel, titled The Believers (Carston 2010Carston, Robyn 2010 “Metaphor: Ad Hoc Concepts, Literal Meaning and Mental Images.” Proceedings of the Aristotelian Society 110: 295–321. ). Therein depression and the unhappiness it is associated with are depicted as a toad or some sort of amphibious creature with features that interfere with everyday human experience.

The literal content of the passage in (3) is taken to take over the metaphorical interpretation process. The set of literal descriptions uttered make up the theme of the metaphor (pervading the text), which, in (3), pertains to the disturbing experience of depression: depression is conceived of as an annoying toad.

To continue, the description of the physical appearance and behavior of a toad adduces propositional (encyclopedic) information about the entity described, i.e. toad. This information, in turn, supplies the contextual assumptions based on which the reader derives implicatures that apply to the entity viewed metaphorically, i.e. depression. As is explained, “from this set of descriptions […], taken as a whole, we derive implications that can plausibly apply to the human experiences of dull depression and raw unhappiness” (Carston 2010Carston, Robyn 2010 “Metaphor: Ad Hoc Concepts, Literal Meaning and Mental Images.” Proceedings of the Aristotelian Society 110: 295–321. , 308). In short, the literal description serves as a source of contextual information from which the reader derives implicatures supporting metaphorical interpretation. To put it plainly, creative metaphor understanding becomes entirely a matter of deriving implicatures: implicated conclusions or output propositions.

Conventional and creative metaphors, then, differ in terms of the types of contents that their interpretation hinges on. Whereas conventional metaphors are understood via explicature (the ad hoc concept), creative ones are understood via implicatures.

Additionally, in keeping with Davidson’s (1984Davidson, Donald 1984 Inquiries into Truth and Interpretation. Oxford: Oxford University Press., 262) position that much of what a metaphor makes us notice is imagistic and, thus, non-propositional in nature, the relevance-theoretic account predicts that a metaphor vehicle, such as the one in (3), constitutes a coherent piece of text, which, apart from generating contextual assumptions, also triggers imagistic effects: mental images of an amphibious creature, as portrayed above. To return to and complete Carston’s remarks (offered above), “from this set of descriptions (and accompanying imagery), taken as a whole, we derive implications that can plausibly apply to the human experiences of dull depression and raw unhappiness” (Carston 2010Carston, Robyn 2010 “Metaphor: Ad Hoc Concepts, Literal Meaning and Mental Images.” Proceedings of the Aristotelian Society 110: 295–321. , 308).

In this respect, creative metaphor interpretation seems to involve the standard inferential process that applies to the derivation of implicated propositional conclusions (or implicatures) from implicated propositional premises, but also allows for the further process of mental image evocation. To the extent that these two processes are distinguishable, it is worth examining whether deriving mental images (or mental image effects) is intrinsically involved in (or essential to) creative metaphor interpretation, a view adopted by Gibbs and Bogdonovich (1999Gibbs, Raymond, and Jody Bogdonovich 1999 “Mental Imagery in Interpreting Poetic Metaphor.” Metaphor and Symbol 14: 37–44. , 43), or not, a view taken by Carston. Let us elaborate on this issue.

Following Colston (2015)Colston, Herbert 2015 Using Figurative Language. Cambridge: Cambridge University Press. , Carston suggests that mental images are “an epiphenomenon or incidental side-effect, with all the work of understanding being done by standard pragmatic processes of inference on conceptual/propositional representations” (Carston 2018Carston, Robyn 2018 “Figurative Language, Mental Imagery and Pragmatics.” Metaphor and Symbol 33: 198–217. , 27–28). Given that the issue of the essentiality of mental images cannot be determined regardless of whether mental images are an essential part of the meaning communicated or intended by the speaker, Carston’s conclusion also suggests that mental images are not communicated. Following Sperber and Wilson (2015)Sperber, Dan, and Deirdre Wilson 2015 “Beyond Speaker’s Meaning”. Croatian Journal of Philosophy 15: 117–149., she notes that “what is overtly communicated is something propositional (truth-evaluable) and thus [images (or their contents) are excluded] from falling in the domain of overt communication” (Carston 2018Carston, Robyn 2018 “Figurative Language, Mental Imagery and Pragmatics.” Metaphor and Symbol 33: 198–217. , 219). On the relevance-theoretic analysis, then, it is implicit content, rather than mental image effects, that is directly or intentionally communicated by the writer. As is pointed out in subsequent research, unlike (weakly) communicated propositions, mental images are not part of the intended import of an ostensive act (Wilson and Carston 2019Wilson, Deirdre, and Robyn Carston 2019 “Pragmatics and the Challenge of ‘Non-Propositional’ Effects.” Journal of Pragmatics 57: 125–148. , 11).

These observations, however, do not imply that mental imagery fulfills a trivial role in the metaphor interpretation process, or more generally, in pragmatic theory. It plays a subsidiary but, still, “significant tangential role” (Carston 2018Carston, Robyn 2018 “Figurative Language, Mental Imagery and Pragmatics.” Metaphor and Symbol 33: 198–217. , 39).

Although it is not specified where mental images are located in relation to the reader’s context, it seems that they reside in her memory and that they interact with contextual assumptions to the benefit of making (more) manifest the implicatures communicated or intended by the speaker/writer in the utterance interpretation process. As is pointed out,

[m]y suggestion is that some verbal metaphors […] can achieve something similar [make propositions manifest] by activating mental images and sustaining them above the threshold of consciousness. These images increase the degree to which certain thoughts/propositions are manifest to readers/hearers, propositions which may be used in deriving (weakly communicated) implications, which contribute to the relevance of the utterance/text.(Carston 2018Carston, Robyn 2018 “Figurative Language, Mental Imagery and Pragmatics.” Metaphor and Symbol 33: 198–217. , 35)

2.2Affective effects

The prospect of associating poetic effects with affective effects is initially entertained by Sperber and Wilson (1995Sperber, Dan, and Deirdre Wilson 1995 Relevance: Communication and Cognition. Oxford: Blackwell., 217–224). In fact, current experimental research in the field of non-propositional effects and the role they play in creative metaphor interpretation indicates that it is not only the derivation of mental images but also that of emotional (affective) effects that partakes of the interpretation process (Ifantidou 2021Ifantidou, Elly 2021 “Non-Propositional Effects in Verbal Communication: The Case of Metaphor.” Journal of Pragmatics 181: 6–16. ). Against this backdrop, an equally important issue also examined pertains to the determination of the kind of link that may hold between mental images and emotive experience.

To elaborate, in light of the fact that language is not emotionally dry and, in this respect, utterance meaning is not a purely cognitive content, Ifantidou’s (2021)Ifantidou, Elly 2021 “Non-Propositional Effects in Verbal Communication: The Case of Metaphor.” Journal of Pragmatics 181: 6–16. experiment is intended to present statistically significant results in favor of the suggestion that interpreting creative metaphors is clearly affected, not only by mental imagery evocation but also by affective responses, and that, in this sense, these two types of cognitive experience play a more active role in the process of metaphorical utterance interpretation than originally acknowledged. The two types of experience appear to interact, forming a non-propositional kind of ground that supports the inferential (metaphorical) processing, resulting in the derivation of novel implicatures. This suggestion is compatible with the view that verbal metaphors display the trait of ineffability (Wilson and Carston 2019Wilson, Deirdre, and Robyn Carston 2019 “Pragmatics and the Challenge of ‘Non-Propositional’ Effects.” Journal of Pragmatics 57: 125–148. ), as well as the view of descriptively ineffable expressions (Blakemore 2011Blakemore, Diane 2011 “On the Descriptive Ineffability of Expressive Meaning.” Journal of Pragmatics 43: 3537–3550. ; de Saussure and Wharton 2020de Saussure, Louis, and Tim Wharton 2020 “Relevance, Effects and Affect.” International Review of Pragmatics 12: 183–205. ).

Importantly, it is concluded that non-propositional effects play a dynamic role in the meaning-making process. Particularly, it is not only the case that interpretation may cause non-propositional effects; at the same time, being imbued with mental images and emotive effects, memorized experiences may cause interpretation.

The results of this experimental research are enlightening, as, ultimately, they do not rule out the prospect that mental images/affective effects are dissociable from the creative metaphor interpretation, despite the case made for the significance of the roles of mental imaging and affective effects in the process of interpretation. As is observed, “how closely linked emotions and mental images are to propositional meaning [inferences or implicatures] derived would require a closer, qualitative evaluation of the data” (Ifantidou 2021Ifantidou, Elly 2021 “Non-Propositional Effects in Verbal Communication: The Case of Metaphor.” Journal of Pragmatics 181: 6–16. , 15).

The experiment to be discussed in the present study is meant to make a contribution to this end, exploring the possibility that mental image/affective effects have an indispensable impact on creative metaphor interpretation. However, this discussion is reserved for Section 4. The section that follows offers an application of the notion of non-propositional effects to the analysis of two poems, which differ as to the sort of image content they carry. These poems also constitute part of the materials utilized for the purpose of the experiment.

4.1Aim

The present experiment aims to help resolve the dilemma raised, initially, about mental imaging and, subsequently, about affective effects: are mental image/emotive effect derivation entailed by (essential to) the creative metaphor interpretation process or are they an epiphenomenon, with all the work of understanding being done by standard pragmatic processes of inference on conceptual/propositional representations? As has been suggested, taking up the issue of the essentiality of non-propositional effects in determining metaphorical interpretation involves addressing the question whether non-propositional effects are communicated by the writer.

The point of departure in the present investigation is to decide on the kind of experimental grounds that can support the fulfillment of the aim defined above. In broad terms, the suggestion that non-propositional effects are part of communicated metaphorical meaning (or play a necessary or essential part in supporting metaphorical interpretation) raises the expectation that readers of a poem who reach the same interpretation of an extended, creative case of metaphor also share the same (or, more accurately, similarly described) non-propositional effects. It is worth noting here that, in examining the extent to which non-propositional effects determine interpretation, it would make little sense to take into consideration cases of a uniform mental image/emotive experience among participants who reached different interpretations.

On this stipulation, it seems fairly obvious that the questions raised above can best be approached subject to answering one further question: to what extent are non-propositional effects constant (or relatively constant) across converging interpreters in the course of interpreting creative metaphor?

The type of experimental frame, then, that suggests itself involves the participation of a number of readers in the interpretation of a specific literary text, such as a poem. Moreover, it involves determining: (a) the rate of a uniform metaphorical interpretation, (b) the rate of a uniform experience of the foregoing non-propositional effects and, finally, (c) comparing (a) and (b) so as to pin down the extent to which they match.

4.2Participants

The participants in the experiment were 180 eighteen-year-old, first-year Greek students of English Language and Literature in Aristotle University (mostly female). The students’ first language is Modern Greek, English being a foreign/second language. The students’ level of competence in English is high (C1/C2 level).

4.3Materials – procedure

The literary texts used in the experiment were the two poems discussed in Section 3 (in Modern Greek). A question worth raising here regards the decision to employ two poems, rather than one; one being highly imagistic, as was noted earlier, the other less so. This decision makes better sense considering the complexity of the issue explored. To be more accurate, while the study of a single poem can provide definite feedback on the type or amount of mental images evoked or the type of emotive response to the text, the results might be considered inconclusive unless the experiment also takes account of the pattern of descriptive content distribution that is employed. For instance, does the literal content of the entire text revolve around the coherent description of an obvious item or subject, or not? This choice of pattern represents a strategic writing choice, which is meant to either augment or attenuate the process of image evocation. A reasonable move, then, would be to contrast the interpretive behavior of two groups of participants engaging in the reading of two poems with differing patterns of descriptive content.

Following the reading activity, participants were given a questionnaire in English comprising the following five questions (i-v). Questions (i), (iii) and (v) required open-ended answers (again in English), so that the students could decide freely on the length of the answer.

1. What is your interpretation of the poem?

2. While reading the poem, did it create images in your mind? Circle one of the answers below: Yes/No.

3. If it did, what were these images?

4. While reading the poem, did you experience any emotions? Circle one of the answers below: Yes/No.

5. If you did, what were these emotions?

The experiment was conducted in a quiet classroom facilitating concentration on the task. The students were separated into two smaller groups of ninety students, group A and group B. The two poems, ‘Candles’ and ‘The Greatest Paradox’, were assigned to groups A and B, respectively. Moreover, students were encouraged to read and interpret the poem that they were assigned.

4.4Methods – rationale

4.5Results and preliminary discussion

Tables 2–4 offer a descriptive summary of the results obtained from the experiment. The figures offered represent numbers of responses collected from the participants, rather than numbers of participants. These figures are raw data followed by percentages.

To start with, Table 2 reflects the extent to which the participants reported experiencing mental images or emotions in the reading process. All 180 responses (ninety responses from each group) reported mental image/emotive experience.77.Access to the participants’ answers can be provided upon request.

Table 3 shows the rate at which the responses collected agree on the same (or similar) interpretations on the part of the participants in both groups.

Table 4 illustrates the rate at which the responses collected converge on the same image content or emotive experience. As suggested in Section 4.1, the rate at which non-propositional effects are experienced uniformly can be determined in light of interpretive consensus. Thus, the rates illustrated below were determined against the entire set of interpretations offered by group A and a subset of interpretations offered by group B (77.8%) (Table 3).

Let us first focus on the results obtained from group A. Regarding Table 3, each response captured a phrase/sentence-long answer to question (i) (Section 4.3). The answers unanimously converged on a singular theme or metaphorical interpretation. They also followed a roughly undifferentiated phraseological pattern of the following sort: the lighted candles represent “the remaining days of our lives; life; stages in life that one goes through; the future”. Conversely, extinguished candles represent “the past; bygone days of one’s youth”. Taking account of this high degree of interpretive agreement, it is no coincidence that the totality of participants also reported experiencing mental image evocation in the reading process, as is shown in Table 2 (a).

A closer look at the results concerning image evocation leads to an additional point of interest. All ninety responses reported the evocation of a specific mental image, described univocally in terms of “lighted candles illuminate a dark place” [Table 4 (a)]. Table 5 (below) details the relevant results. The figure provided in the second column represents the frequency with which the wording used by the ninety participants (in order to describe their experience of mental imaging) occurs in their reports. A total of ninety responses were collected.88.This involves a correspondence of one (student)-to-one (response), in the sense that each student offered one response. On most of the data-collecting occasions that follow, each student provided two–three responses on average.

Equally remarkable is the result pertaining to emotive responses. Once again, all ninety participants reported experiencing emotions in the course of interpretation [Table 2 (b)]. However, while affective experience was common to all readers, not all participants experienced identical emotions, as can be observed [Table 4 (b)]. Table 6 (below) elaborates on the relevant results. The figures provided in the third column represent the frequency with which emotion terms were reported. A total of 180 terms were collected.

The majority (77.8%) of the answers brought up two emotion terms that can be viewed as linked to a single family, i.e. “optimism”, whereas the remaining (22.2% of the) answers varied greatly from phrases linked to the domain of “cheerfulness” to other phrases linked to various domains: “horror, longing, nervousness, lust, sadness”.

Moving on to group B, each response to question (i) framed a phrase/sentence-long answer, as in the case of group A. 77.8% of responses (or seventy out of ninety) appear to agree on a single thematic or metaphorical interpretation of the poem (Table 3). Considering the relevance-theoretic claim that communication involves a risk of misinterpretation, it is reasonable to assume that not all readers might reach the same interpretation of a poetic text. However, this poses no challenge to the present experimental work, since, as was pointed out (Section 4.1), in order to determine the essentiality of non-propositional effects in metaphorical interpretation, we aim to measure the extent of uniformity of mental image/emotive effect experience against the extent of interpretive consensus reached by the participants (rather than against the sum of the participants).

The interpretive consensus reached by this subgroup (77.8%) of responses can be summarized, albeit rather loosely, as follows, on the basis of word-choice overlap: the poem is about “people having no control over their lives/helplessness”. The answers provided by the remaining students (22.2%) adopt a different phraseology, which does not allow identification of an obvious link to the above-mentioned interpretation. Nor do they readily converge on a single interpretation, or an interpretation that can be coherently generalized: “hatred in the world; war of nations/international relationships; feeling of not belonging; giving up on your dreams; issue of conflict of all kinds; people experience internal struggles; people realize mistakes made in the past; life experienced in a disenchanting world”.

Given the absence of a full consensus as to the thematic interpretation of the poem, a point of interest is that all members of group B experienced mental imaging [Table 2 (a)]. Yet, the image content visualized varied with the experiencer [Table 4 (a)]. Table 7 (below) refines the related results by presenting the frequency with which the terms employed appear in the participants’ reports. Α total of 142 descriptive phrases were collected from the subgroup of students who reached a consensus on the interpretation of the poem. A large number of them were reported even though they do not occur in the poem.

More concretely, the highest rate of common visualization experience was 20%, as only twenty-eight of the total of 142 answers collected described the visualization of the same mental image content: naval combat. The rest of the answers (114 out of 142) report a diversity of images that makes up the remaining 80% of responses. These answers include the following collection of expressions: fire/flames, emotionless person, lack of motion, huge dark world, nature, monster, statue, soldier, paradise, canons, sinking, drowning people, smoke, water. Importantly, the content words reported by group B are not mentioned in the poetic text, while key content words (lighted candles) used by group A are also used in the text.

Moving on to the aspect of emotive experience, once again, all members in group B reported experiencing affective effects [Table 2 (b)]. However, not all members of the subgroup of participants who agreed on the interpretation of the poem reported experiencing the same affective effect [Table 4 (b)]. Table 8 details the results shown in Table 4 (b). In particular, it presents the frequency with which the terms used by the seventy participants (to express their emotive experience) occur in their reports. A total of 220 terms were collected.

As becomes evident, the highest rate of common emotive experience was 24.5%. Fifty-four of the total of 220 responses that were gathered reported the experience of the same emotion, that of horror. The remaining answers (75.5%) report a range of emotions that splits the participants’ emotive experience into numerous distinct experiences: sadness, exasperation, irritability, rage, surprise, nervousness, suffering, neglect. Noteworthy is the fact that 18.3% of terms mentioned represent mental states that do not qualify as emotion concepts, according to Shaver et al.’s (1987)Shaver, Philip, Judith Schwartz, Donald Kirson, and Cary O’Connor 1987 “Emotion Knowledge: Further Exploration of a Prototype Approach.” Journal of Personality and Social Psychology 52: 1061–1086. typology and after inter-annotator assessment.

In order to measure the statistical significance of the difference between the two groups with respect to mental image evocation and affective experience, a t-test was used so as to compare the means of the two populations (groups). Judging from the results (Table 9), we can observe that the difference is statistically significant in both respects, the level of significance reaching 95%, given that p (sig2-tailed) is smaller than 0.05.

4.6Interpretation of results

As mentioned previously (Section 4.1), the present experiment was carried out with a view to determining whether the derivation of mental image/emotive effects constitutes an essential factor in the creative metaphor interpretation process. If it is, then these effects are part of the writer’s intention to communicate (metaphorical) meaning or, in other words, they are part of an ostensive act.

A possible objection concerns the premises of the present analysis: readers of creative metaphors who reach a common metaphorical interpretation also share the same (or similarly described) non-propositional effects. In other words, while not all of the research participants who reached the same interpretation shared the same mental image/emotive experience, this fact does not rule out the possibility that mental imagery or emotions determine interpretation: it could simply mean that the stimuli are too vague or ambiguous, and that they have all been understood, but in different ways (similarly to weak implicatures).

The plausibility of this objection notwithstanding, the present findings contribute further evidence for the claim that non-propositional effects are not part of the (metaphorical) content communicated. In fact, they corroborate the original expectation that mental images, as divergent as, say, ‘statue’ and ‘monster’, cannot be considered necessary (essential) conditions for the determination of a specific metaphorical interpretation, e.g. that of helplessness (poem B), at least as compared to the (more) determinate utterance meaning, intended by the communicator. Nor can the experience of an emotion such as rage be deemed a necessary condition for the delivery of the foregoing interpretation, if a host of emotions were experienced by a number of readers, who were otherwise exposed to the same linguistic stimulus in the interest of delivering the same interpretation.

To continue, while (weak) implicatures take a propositional form, mental images/affective effects do not. Unlike non-propositional effects, propositions are known to bear the following properties: they are true/false and, also, they can be related, so that a proposition can be deduced from another in a deductive process. In this respect, from a relevance-theoretic point of view, (weak) implicatures constitute contextualizable assumptions, in other words assumptions that can combine with existing contextual information in the interest of delivering a relevant (metaphorically intended) interpretation. This contextualizability does not appear to be a property of non-propositional effects: contextual assumptions are neither deducible from mental images/emotions, nor vice versa. Nor can mental images or emotions be deduced from other mental images or emotions. As may be recalled, according to the findings of the present experiment, the reading process triggered a wide range of emotive effects/mental images, none of which can be seen as deducible from the context supplied by the text or from further non-propositional effects. Overall, non-propositional effects may concur with the context, or occur in parallel with the contextualization process (as noted in Sections 3–4) but they do not appear to be internal to it.

Moving on to the results of the experiment, a higher level of interpretive uniformity was achieved among participants in group A. This observation is ascribable to the poem’s focus on the description of a single object/item, i.e. candles. This focus secures an amount of coherent descriptive content facilitating context selection. In this sense, the relatively lower rate of interpretive uniformity is unsurprising in group B, where context selection decelerates in light of the lack of (obvious) focus or coherent descriptive content.

Regarding the issue of mental images, a striking finding is that all readers, across groups, evoked mental images. Based on this finding, it would be tempting to conclude, contra Carston, that mental imagery is instrumental to the creative metaphor interpretation and, thus, that no creative metaphor interpretation is deliverable without support from mental image evocation. However, this is a conclusion that calls for finer observation, especially in light of the finding that, unlike the situation in group A, where all reports converge on the same interpretation and the same mental image effects, this particular result does not obtain for group B, where, in fact, only 20% of the reports agree on the visualization of the same image content.

These finer observations call explicit attention to a distinction that is typically implied in the relevant literature: mental imaging and mental image content. The former yields the latter. Equating these two aspects of the mental image derivation process might prove fallacious in our endeavor to determine whether ‘mental images’ are ostensively communicated or play an essential role in creative metaphor interpretation. In actual fact, the question whether mental images are communicated is an abbreviation of the following longer question: can specific mental image content/effects be communicated by an utterance (or textual input)?

As it turns out, while it can be argued that the cognitive operation of mental imagery is constant among creative metaphor interpreters, the derivation of a particular mental image effect is not, in that our cognitive ability to visualize linguistic input cannot warrant the derivation of any particular amount or quality of image content. Therefore, a more plausible view would be that (specific) mental image effects derive, not as an integral part of retrieving the writer’s communicative intention but, rather, as a collateral gain achieved by the reader, of her own accord (or independently of the writer’s intention to communicate information); and, while we cannot overlook the constant operation of a mental imaging mechanism whose task is to enhance the reader’s experience of metaphorical interpretation with illustrated content, this operation does not secure the derivation of a specific, writer-intended image effect but, rather, encourages the derivation of any set of mental image effects. To a large extent, these effects vary with the reader and, more accurately, they vary with the context the reader supplies upon processing the descriptive content of the text.

It follows that creative metaphor interpretation (or misinterpretation) does not depend on the retrieval of mental images residing in the writer’s mind. Rather, in more familiar relevance-theoretic terms, it relies on the contextualization of the poet’s utterances in the reader’s context, independently of the kind of mental image content that may be evoked. On this suggestion, mental image effects may serve to interact with and illustrate the set of contextual assumptions that will contextualize the poet’s utterances as part of the interpretation process. This sort of “illustrated”, as it were, contextualization can be said to yield further non-propositional effects. Moreover, it can be said to increase the relevance of the interpretation, in case the mental illustration is coherent with the interpretation.

It transpires, then, that any considerations of the role of writer (metaphorical) intentions in the course of communication need to be viewed on a level other than the ostensive-inferential one. In particular, the writer’s intention lies not in communicating mental image effects to the reader but in manipulating the cognitive operation of mental imaging, conceivably to the benefit of granting the reader instant access to his perspective. For instance, the greater the amount of coherent descriptive density employed, the greater the prospect of creating specific image content across readers.

It also seems reasonable that the differentiation between a cognitive mechanism and the non-propositional effects that it may yield carries over to the issue of affective effects, as is testified by the respective results. Whereas all participants (across groups) reported experiencing emotions, not all of them noted the same emotion terms, or, more generally, terms linked to the same emotion domain, as becomes especially evident in group B. Since emotive experience was common to all participants, it stands to reason that the mechanism of affective effect derivation plays a constant role in interpreting creative metaphors. However, it does not follow that metaphorical interpretation depends on the derivation of specific affective effects. In fact, the result showing non-uniform emotion reports in group A, despite 100% consensus on the interpretation, seems to be a clear argument to support the claim that affective effects do not drive interpretation’. Thus, metaphorical interpretation depends on the selection of the context in which metaphorically intended utterances will be contextualized. This context selection will occur independently of the affective effects that may accompany the implicatures derived (metaphorical interpretation). In this line of argument, the derivation of a specific affective effect may add an emotive element to the reader’s context, leading to an emotionally laden metaphorical interpretation, which can, once again, serve to grant the reader direct access to the writer’s perspective or psychological disposition.

In light of all this, a more plausible and presumably less strong view is the one held by Ifantidou (2021)Ifantidou, Elly 2021 “Non-Propositional Effects in Verbal Communication: The Case of Metaphor.” Journal of Pragmatics 181: 6–16. . Non-propositional effects play a significant role in the meaning-making process. On this account, they can be viewed as being caused by interpretation but also as the cause of interpretation (also see Wharton and Strey 2019Wharton, Tim, and Claudia Strey 2019 “Slave to the Passions: Making Emotions Relevant.” In Relevance, Pragmatics and Interpretation, ed. by Robyn Carston, Billy Clark, and Kate Scott, 253–266. Cambridge: Cambridge University Press. ). We take it that this claim does not imply, though, that non-propositional effects determine interpretation; while they might occasionally motivate it, ultimately, interpretation can be achieved without them.