Pith. sign in

REVIEW 4 major objections 6 minor 51 references

Meanings are like Onions: a Layered Approach to Metaphor Processing

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper proposes a three-layered, onion-like model of metaphor meaning—context, conceptual blending, and pragmatic intention—as the basis for computational metaphor processing.

desk verdict A well-written theoretical proposal for a three-layer metaphor annotation schema whose pragmatic layer is genuinely new, but the abstract oversells it as a formal framework and the annotatability of pragmatic features is entirely untested. read the letter →

arxiv 2507.10354 v2 pith:WB7Y3TH2 submitted 2025-07-14 cs.CL

Silvia Cappa , Anna Sofia Lippolis , Stefano Zoia This is my paper · ORCID
classification cs.CL
keywords metaphorrepresentationpragmaticsconceptualblendingspeechacttheorycomputationalprocessingmultilayeredmeaningannotationframeworkAI
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Metaphorical meaning, this paper argues, is not a flat source-to-target mapping but a multi-layered phenomenon that computational systems should represent as an onion. The proposed structure has three layers: an outer layer of contextual and annotator metadata, a middle layer in which source and target concepts are blended into an emergent meaning, and an inner layer of pragmatic intention that treats a metaphor as a speech act—what it is meant to do, such as warn, request, or persuade. The paper claims that unifying these layers into one framework lets computational metaphor processing capture both what metaphors map and what they accomplish in communication. If the model is adopted, metaphor datasets would include attitude, illocutionary force, perlocutionary effects, and visual tone alongside traditional conceptual annotations, and systems would generate and interpret metaphors with context sensitivity rather than surface association.

What carries the argument

The central mechanism is the onion representation itself: a stack of three nested layers of annotation for a single metaphorical object. Layer 1 holds contextual metadata (domain, provenance, frame links, annotator demographics). Layer 2 holds the conceptual blend: the source and target as blendable input spaces, the blending principle that justifies their combination, and the blended concept that emerges. Layer 3 holds a pragmatic description in terms of illocutionary act (including directive kind along an assertiveness continuum), perlocutionary effect (emotions and an efficacy rating), and pragmatic/visual clues. Each inner layer is interpreted through the outer ones, converting a metaphor into a structured object that ties frame-level knowledge to communicative function, and the paper proposes that this stack be implemented by combining a prototypical conceptual-combination system with a logic-augmented generation pipeline.

What would settle it

Run an annotation study in which independent annotators label a fixed set of metaphorical images with the proposed pragmatic categories, and measure inter-annotator agreement (for example with Cohen's kappa on the illocutionary act and directive kind, and with intraclass correlation on efficacy). If agreement is not substantially above chance, the pragmatic layer cannot serve as a reliable annotation target, and the framework's promise of grounding new datasets and evaluations would be undermined.

Watch

Extended reading notes

Core claim

The paper's central proposal is a three-dimensional 'onion' model of metaphor representation. The outermost layer records the communicative object's context: domain, provenance, links to knowledge bases, and annotator background, acknowledging that interpretation is culturally situated. The middle layer represents the conceptual combination of source and target, encoding blendables, the blending principle that licenses the mapping, and the resulting blended space. The innermost layer introduces a pragmatic vocabulary—attitude, illocutionary act with directive kind, perlocutionary effect, efficacy, visual clues, and tone of voice—so that a metaphorical utterance can be analyzed as an action that produces effects in listeners. The paper argues that uniting cognitive and pragmatic analysis in this single layered framework is the missing connection between computational metaphor research and pragmatics, and that it lays the groundwork for richer datasets, standardized evaluation, and systems that reason about figurative meaning beyond surface associations.

Load-bearing premise

The framework depends on the assumption that people can reliably agree, using the proposed categories, on pragmatic features of a metaphorical image—such as its illocutionary act, directive kind, perlocutionary effect, and efficacy—but the paper provides only a hand-written example, with no annotation manual, no agreement statistics, and no dataset that currently annotates these dimensions.

Editorial extensions

If this is right

  • Metaphor datasets could be constructed with annotations for attitude, speech acts, perlocutionary effects, and visual tone in addition to source-target mappings, yielding the first gold-standard pragmatic metaphor resources.
  • Neuro-symbolic systems could combine prototypical conceptual-combination output with logic-augmented knowledge-graph generation to produce representations that include speaker intent and emotional effect.
  • Metaphor generation systems, especially in chatbots and digital companions, could choose metaphors by pragmatic fit so that the output is functionally appropriate to the conversational context.
  • Metaphor understanding systems could filter interpretations through pragmatic features such as speaker goal, emotional tone, and discourse situation, improving disambiguation in dialogue, narratives, and scientific text.
  • Metaphor-aware applications in education, science communication, and doctor-patient interaction could exploit the same layered representation to align figurative language with audience sensitivity.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If adopted, evaluation targets may shift from metaphor detection accuracy toward perlocutionary success—whether a warning metaphor actually alarms or persuades—since the framework makes the effect on the listener an explicit annotation target.
  • The paper's remark that meaning can be represented as a spectrum suggests a testable extension: modeling directive-kind judgments (from gentle suggestion to command) as a continuous regression, and checking whether the predicted position correlates with annotators' efficacy ratings.
  • The onion stack could become dynamic: if metaphor reinterpretation and resistance are treated as updates to the pragmatic layer across conversational turns, the framework would connect to social and inter-bodily accounts of metaphor and allow formal tracking of when a community accepts or rejects a metaphor.
  • Because Layer 1 explicitly records annotator cultural background, aggregating pragmatic labels as a distribution over diverse annotators—rather than collapsing to a single gold label—could turn the acknowledged subjectivity of metaphor interpretation into a measurable signal, consistent with the paper's call for diverse annotator pools.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. This position paper proposes a three-layered model of metaphorical meaning, analogized to an onion: an outer contextual/annotator-metadata layer, a middle conceptual blending layer encoding source-target mappings and emergent blends, and an inner pragmatic layer capturing attitude, illocutionary act, directive kind, perlocutionary effect, efficacy, and visual clues/tone. The authors argue that such a stratified representation, grounded in CMT, CBT, and speech act theory, can unify conceptual and pragmatic aspects of metaphor and thereby support future computational metaphor processing, dataset creation, and evaluation. The paper is explicitly theoretical: it contains no experiments, no formal ontology, and no implemented system, instead offering examples on anti-smoking images and recommending existing tools (METCL, LAG, Emotion Frame Ontology) as candidate implementations. It concludes by outlining implications for datasets, knowledge representation, metaphor generation/understanding, and socially beneficial applications.

Significance. If the framework is adopted and operationalized, it would be a valuable contribution that connects cognitive metaphor theory with computational pragmatics, addressing a real gap: most metaphor datasets and systems ignore illocutionary and perlocutionary dimensions. The paper is strong in its theoretical grounding, its use of existing ontological resources (Blending Ontology, MetaNet, EFO), and its clear worked examples for Layers 1 and 2. It also explicitly names candidate neurosymbolic implementations, which is useful for reproducibility. However, the significance is conditional on the operationalizability of Layer 3, the pragmatic layer, for which no annotation reliability evidence is provided. The paper is a credible research proposal, but its central promise to 'lay the groundwork for new datasets and evaluation' is not yet supported by evidence that the proposed pragmatic categories can be applied consistently by human annotators.

major comments (4)
  1. [§4 and §3.1.3] The entire framework rests on the assumption that the Layer 3 categories—illocutionary act, directive kind on an assertiveness continuum, perlocutionary effect, efficacy, visual clues, and tone—can be annotated reliably on images such as Figure 1b. Section 3.1.3 provides exactly one hand-authored example (Figure 4) and no annotation manual, rubric, or inter-annotator agreement data. Section 4 concedes that 'no single metaphor-related dataset annotates attitude, speech acts and clues of utterances' and that a gold-standard pragmatic dataset is still needed. Because the paper's stated contribution is to ground future datasets and evaluations, the untested annotatability of Layer 3 is a load-bearing empirical premise. The paper should either report a small pilot annotation study demonstrating feasibility or explicitly state that annotation reliability is an open empirical question and soften the claim that the framework already lays the groundwork for such datasets.
  2. [Abstract and §3.1] The abstract and Section 3.1 describe the model as a 'single formal framework', but the paper contains no formal definitions, no formal syntax or semantics, and no ontology specification; the layers are described in prose and figures, and Section 3.1 itself says the framework 'serves as a vocabulary'. This is a mismatch between the stated contribution and the actual content. If 'formal' is intended literally, the missing formalization should be provided or a pointer to a future formalization given; if it is used informally, the term should be replaced throughout to avoid overclaiming.
  3. [§3.1.1, §3.1.2, §3.1.3] The layer boundaries are described inconsistently. The abstract received with this manuscript states Layer 1 concerns 'content analysis' and 'basic conceptual elements', while the full text says Layer 1 is 'contextual information' and calls it the 'external context layer'; the first sentence of §3.1.1 then calls it the 'conceptual content layer'. Layer 2 is described both as the 'cognitive context of the element' and as the source-target blending analysis. These inconsistencies undermine the central claim that the layers are separable and can be annotated independently. A precise fixed definition of each layer, with explicit inclusion/exclusion criteria, is needed before the framework can serve as a basis for annotation.
  4. [§3.1.3] The definitions of Perlocutionary Effect and Efficacy conflate the annotator's subjective response with properties of the utterance/image. Perlocutionary Effect asks annotators to report 'the emotions evoked by the metaphor in the annotator', while Efficacy asks for a judgment of effectiveness 'relative to the presumed illocutionary intention'. This confounds reception (what the annotator feels) with inference (what the speaker intended), and the paper itself acknowledges in §2.5 that illocutionary forces are often blended and context-dependent. The scheme would be more defensible if it distinguished 'intended perlocutionary effect' from 'actual emotional response', or if it specified a procedure for deriving the latter from a shared reading of the communicative situation.
minor comments (6)
  1. [§3.1] There is a typo 'framrwork' in the opening sentence of §3.1; the intended word is 'framework'.
  2. [§2] The heading of §2 reads 'theorical background'; this should be 'theoretical background'.
  3. [§3.1.2] The phrase 'yeld the mappings' contains a typo; it should be 'yield the mappings'.
  4. [§3.1.2 and elsewhere] The system name 'METCL' is inconsistently spaced as 'MET CL' in several places (e.g., §3.1.2, Figure 3 caption, §4). Please standardize the notation.
  5. [§3.1.3] In the bullet list for Perlocutionary Act, the definition of perlocutionary act is given in parentheses in §2.5 as 'the name for the speech act performed by saying something', which is inaccurate: in Austin's terminology, a perlocutionary act is the act performed by means of saying something, not the name for the speech act. The paper should cite a standard definition or rephrase to avoid confusion.
  6. [§5] The conclusion claims the framework 'can be used' in several applications, but no concrete evaluation metrics or falsifiable predictions are given; consider adding an explicit list of testable predictions to make the proposal more actionable.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation: the paper is a theoretical proposal with no fitted parameters or predictions; the only self-citations are illustrative tool suggestions that are not load-bearing.

full rationale

This is an explicitly theoretical position paper. It contains no equations, no fitted parameters, and no empirical predictions, so there is no derivation chain in which an output could reduce to an input. The central proposal is that metaphorical meaning can be usefully analyzed in three layers (context, conceptual combination, pragmatics); this is presented as a vocabulary for future datasets and systems, not as a result derived from prior work. The paper's own Section 5 states: 'The main current limitation of this framework is the lack of computational implementation,' confirming that no computational claim is being made or validated. The only self-referential elements are citations to the authors' own tools: METCL [35], whose author list includes coauthor Zoia, is suggested in Section 3.1.2 as a possible mechanism for Layer 2 ('Our proposal for Layer 2 is to use a conceptual combination system like METCL as the reasoning mechanism'), and LAG [25], which includes coauthor Lippolis, is mentioned in Section 4 as potentially integrable with METCL. These citations are illustrative, not load-bearing: the layered framework would stand unchanged if those tools were replaced by any other conceptual-combination or neurosymbolic system. The paper also relies on EFO [51] for emotion categories, but that is an external ontology, not an output of this paper. No uniqueness theorem is imported, no ansatz is smuggled via self-citation, and no empirical pattern is renamed as a prediction. Accordingly, there is no significant circularity, only minor self-citation in the proposed implementation outlook, which is why the score is 2 rather than 0.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The framework makes no fitted predictions, so there are no free parameters. The illustrative numbers, such as an efficacy score of 4/5 in Figure 4, are hand-written examples, not fitted values. No new entities are posited; the pragmatic vocabulary reuses speech act and emotion ontology categories. The axioms are inherited from cognitive linguistics, pragmatics, and the paper's own analytical choice to decompose meaning into separable layers.

assumptions (4)
  • domain assumption CMT and CBT provide adequate accounts of how metaphors create meaning, so source-target mapping and blending can ground Layer 2.
    Invoked throughout Section 2.1 and Section 3.1.2 without new evidence; the framework inherits the validity of these theories.
  • domain assumption Speech act categories from Searle and Austin can be applied to visual metaphors and read off by annotators.
    Section 3.1.3 proposes annotating attitude, illocutionary act, and perlocutionary effect on an image; no inter-annotator reliability evidence is given.
  • ad hoc to paper Meaning can be decomposed into separable layers (context, conceptual, pragmatic) that can be annotated independently.
    Section 3.1 introduces the onion assumption as an analytical choice, not as a demonstrated property of metaphor processing.
  • domain assumption Illocutionary intention can be treated as a property of the utterance rather than a mental state, following Austin.
    Section 2.5 adopts this to avoid opacity of mental states; it is a philosophical commitment, not an empirical result.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Meanings are like Onions: a Layered Approach to Metaphor Processing." pith.science (2026). https://pith.science/paper/WB7Y3TH2

@misc{pith2026250710354,
  author       = {Pith},
  title        = {Pith review of: Meanings are like Onions: a Layered Approach to Metaphor Processing},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WB7Y3TH2}},
  note         = {Machine review of arXiv:2507.10354}
}
read the original abstract

Metaphorical meaning is not a flat mapping between concepts, but a complex cognitive phenomenon that integrates multiple levels of interpretation. In this paper, we propose a stratified model of metaphor processing that treats meaning as an onion: a multi-layered structure comprising (1) content analysis, (2) conceptual blending, and (3) pragmatic intentionality. This three-dimensional framework allows for a richer and more cognitively grounded approach to metaphor interpretation in computational systems. At the first level, metaphors are annotated through basic conceptual elements. At the second level, we model conceptual combinations, linking components to emergent meanings. Finally, at the third level, we introduce a pragmatic vocabulary to capture speaker intent, communicative function, and contextual effects, aligning metaphor understanding with pragmatic theories. By unifying these layers into a single formal framework, our model lays the groundwork for computational methods capable of representing metaphorical meaning beyond surface associations, toward deeper, more context-sensitive reasoning.

Figures

Figures reproduced from arXiv: 2507.10354 by the authors.

Figure 1
Figure 1. Examples of anti-smoking imagery: the same metaphor can be conveyed in different contexts and with different goals. 3.1.1. Layer 1: The external context layer The conceptual content layer includes metadata of the communicative object we are analyzing: domain, provenance, connections to existing knowledge bases, etc. It also includes annotator metadata [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. An example of “Layer 1" analysis for Figure 1b, showing the annotation of its content and related metadata. 3.1.2. Layer 2: The conceptual combination layer This layer includes the cognitive context of the element, and in particular what enables the metaphorical mapping by taking into account only the two domains employed in the metaphors. Analyses that detect source and target concepts in a sentence or image are pa… view at source ↗
Figure 3
Figure 3. An example of “Layer 2" analysis for Figure 1b, showing the result of the conceptual combination performed by METCL . 3.1.3. Layer 3: The pragmatic layer Compared with the first two layers, the third remains largely under-investigated from a computational perspective. Our aim is to discuss a categorization that would allow for its effective implementation, also with a view to the composition and annotation of a data… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: An example of “Layer 3" analysis for Figure 1b, showing its description according to pragmatics. 4. Implications for computational metaphor processing and future work Our proposal holds several implications for computational metaphor processing. In this section, we out…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 47 canonical work pages

  1. [1]

    K. S. Moser, The role of metaphors in acquiring and transmitting knowledge, European perspectives on learning at work: the acquisition of work process knowledge (2004) 148–163

    work page 2004

  2. [2]

    Lakoff, M

    G. Lakoff, M. Johnson, Metaphors We Live By, University of Chicago Press, 1981

    work page 1981

  3. [3]

    R. M. Hicke, R. D. Kristensen-McLachlan, Science is exploration: Computational frontiers for conceptual metaphor theory, arXiv preprint arXiv:2410.08991 (2024)

    work page Pith review arXiv 2024

  4. [4]

    Dancygier, Figurativeness, conceptual metaphor, and blending, in: The Routledge Handbook of metaphor and language, Routledge, 2016, pp

    B. Dancygier, Figurativeness, conceptual metaphor, and blending, in: The Routledge Handbook of metaphor and language, Routledge, 2016, pp. 46–59

    work page 2016

  5. [5]

    Wijesiriwardene, R

    T. Wijesiriwardene, R. Wickramarachchi, B. G. Gajera, S. M. Gowaikar, C. Gupta, A. Chadha, A. N. Reganti, A. Sheth, A. Das, Analogical–a novel benchmark for long text analogy evaluation in large language models, arXiv preprint arXiv:2305.05050 (2023)

    arXiv 2023

  6. [6]

    Nezhurina, L

    M. Nezhurina, L. Cipolina-Kun, M. Cherti, J. Jitsev, Alice in wonderland: Simple tasks showing complete reasoning breakdown in state-of-the-art large language models, arXiv preprint arXiv:2406.02061 (2024)

    arXiv 2024

  7. [7]

    H. P. Grice, Meaning, Readings in the Philosophy of Language (1971) 436–444

    work page 1971

  8. [8]

    Fauconnier, M

    G. Fauconnier, M. Turner, Conceptual blending, form and meaning, Recherches en communication 19 (2003) 57–86

    work page 2003

Show all 51 references
  1. [9]

    K. J. Holyoak, D. Stamenković, Metaphor comprehension: A critical review of theories and evidence., Psychological bulletin 144 (2018) 641

    2018

  2. [10]

    C. J. Fillmore, Frame semantics, in: Cognitive Linguistics: Basic Readings, De Gruyter Mouton, Berlin, New York, 2006, pp. 373–400

    2006

  3. [11]

    T. W. Jensen, The world between us: The social affordances of metaphor in face-to-face interaction, RASK Internationalt tidsskrift for sprog og kommunikation 47 (2018) 45–76

    2018

  4. [12]

    Semino, Metaphor in discourse, Cambridge University Press Cambridge, 2008

    E. Semino, Metaphor in discourse, Cambridge University Press Cambridge, 2008

    2008

  5. [13]

    Titchkosky, Reading and writing disability differently: The textured life of embodiment, University of Toronto Press, 2007

    T. Titchkosky, Reading and writing disability differently: The textured life of embodiment, University of Toronto Press, 2007

    2007

  6. [14]

    Linell, Rethinking language, mind, and world dialogically, IAP, 2009

    P. Linell, Rethinking language, mind, and world dialogically, IAP, 2009

    2009

  7. [15]

    Schalk, Metaphorically speaking: Ableist metaphors in feminist writing, Disability Studies Quarterly 33 (2013)

    S. Schalk, Metaphorically speaking: Ableist metaphors in feminist writing, Disability Studies Quarterly 33 (2013)

    2013

  8. [16]

    R. W. Gibbs, Metaphor wars, Cambridge University Press, 2017

    2017

  9. [17]

    R. W. Gibbs Jr, J. Siman, How we resist metaphors, Language and Cognition 13 (2021) 670–692

    2021

  10. [18]

    Gangemi, M

    A. Gangemi, M. Alam, L. Asprino, V. Presutti, D. R. Recupero, Framester: A wide coverage linguistic linked data hub, in: Knowledge Engineering and Knowledge Management: 20th International Conference, EKAW 2016, Bologna, Italy, November 19-23, 2016, Proceedings 20, Springer, 20...

    2016

  11. [19]

    Gangemi, M

    A. Gangemi, M. Alam, V. Presutti, Amnestic forgery: An ontology of conceptual metaphors, in: S. Borgo, P. Hitzler, O. Kutz (Eds.), Formal Ontology in Information Systems - Proceedings of the 10th International Conference, FOIS 2018, Cape Town, South Africa, 19-21 September 201...

    2018 doi

  12. [20]

    Hamilton, Towards an ontology for propaganda detection in news articles, in: The Semantic Web: ESWC 2021 Satellite Events, 2021, pp

    K. Hamilton, Towards an ontology for propaganda detection in news articles, in: The Semantic Web: ESWC 2021 Satellite Events, 2021, pp. 230–241. URL: https://doi.org/10.1007/978-3-030-80418-3_35. doi: 10.1007/978-3-030-80418-3_ 35

    2021 doi

  13. [21]

    Mitrović, C

    J. Mitrović, C. O’Reilly, M. Mladenović, S. Handschuh, Ontological representations of rhetorical figures for argument mining, Argument & Computation 8 (2017) 267–287. URL: https://doi.org/10.3233/aac-170027. doi: 10.3233/aac- 170027

    2017 doi

  14. [22]

    Krennmayr, G

    T. Krennmayr, G. Steen, Vu amsterdam metaphor corpus, Handbook of linguistic annotation (2017) 1053–1071

    2017

  15. [23]

    Gordon, J

    J. Gordon, J. R. Hobbs, J. May, M. Mohler, F. Morbini, B. Rink, M. Tomlinson, S. Wertheim, A corpus of rich metaphor annotation, in: Proceedings of the Third Workshop on Metaphor in NLP, 2015, pp. 56–66

    2015

  16. [24]

    Dodge, J

    E. Dodge, J. Hong, E. Stickles, O. David, The metanet wiki: A collaborative online resource for metaphor and image schema analysis, in: 12th International Cognitive Linguistics Conference (ICLC 12), 2013

    2013

  17. [25]

    A. S. Lippolis, A. G. Nuzzolese, A. Gangemi, Enhancing multimodal analogical reasoning with logic augmented generation, arXiv preprint arXiv:2504.11190 (2025)

    2025

  18. [26]

    Semino, J

    E. Semino, J. Heywood, M. Short, Methodological problems in the analysis of metaphors in a corpus of conversations about cancer, Journal of pragmatics 36 (2004) 1271–1294

    2004

  19. [27]

    R. Mao, X. Li, H. Kai, M. Ge, E. Cambria, Metapro online:: A computational metaphor processing online system, in: Proceedings of the 61st annual meeting of the association for computational linguistics, Association for Computational Linguistics (ACL), 2023

    2023

  20. [28]

    Y. Tian, N. Xu, W. Mao, A theory guided scaffolding instruction framework for llm-enabled metaphor reasoning, in: Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Pape...

    2024

  21. [29]

    Shutova, Annotation of linguistic and conceptual metaphor, Handbook of linguistic annotation (2017) 1073–1100

    E. Shutova, Annotation of linguistic and conceptual metaphor, Handbook of linguistic annotation (2017) 1073–1100

    2017

  22. [30]

    Z. Rosen, Computationally constructed concepts: A machine learning approach to metaphor interpretation using usage-based construction grammatical cues, in: Proceedings of the workshop on figurative language processing, 2018, pp. 102–109

    2018

  23. [31]

    Wachowiak, D

    L. Wachowiak, D. Gromann, Does gpt-3 grasp metaphors? identifying metaphor mappings with generative language models, in: Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2023, pp. 1018–1032

    2023

  24. [32]

    A. R. Akula, B. Driscoll, P. Narayana, S. Changpinyo, Z. Jia, S. Damle, G. Pruthi, S. Basu, L. Guibas, W. T. Freeman, et al., Metaclue: Towards comprehensive visual metaphors research, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2023,...

    2023

  25. [33]

    Z. Y. Yang, Z. Zhang, Y. Miao, The elco dataset: Bridging emoji and lexical composition, in: Proceedings of the 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation (LREC-COLING 2024), 2024, pp. 15899–15909

    2024

  26. [34]

    Gangemi, A

    A. Gangemi, A. G. Nuzzolese, Logic augmented generation, Journal of Web Semantics 85 (2025) 100859. URL: https: //www.sciencedirect.com/science/article/pii/S1570826824000453. doi:https://doi.org/10.1016/j.websem.2024. 100859

    2025 doi

  27. [35]

    Lieto, G

    A. Lieto, G. L. Pozzato, S. Zoia, The Delta of Thought: Channeling Rivers of Commonsense Knowledge in the Sea of Metaphorical Interpretations, in: Proceedings of IJCAI 2025, 34th International Joint Conference on Artificial Intelligence, AAAI Press, Montréal, Canada, 2025. URL...

    2025

  28. [36]

    Lieto, F

    A. Lieto, F. Perrone, G. L. Pozzato, E. Chiodino, Beyond subgoaling: A dynamic knowledge generation framework for creative problem solving in cognitive architectures, Cognitive Systems Research 58 (2019) 305–316. URL: https: //linkinghub.elsevier.com/retrieve/pii/S138904171930...

    2019 doi

  29. [37]

    H. P. Grice, Logic and conversation, in: D. Davidson (Ed.), The logic of grammar, Dickenson Pub. Co., 1975, pp. 64–75

    1975

  30. [38]

    J. R. Searle, Metaphor, in: A. Ortony (Ed.), Metaphor and Thought, Cambridge University Press, 1993, pp. 83–111

    1993

  31. [39]

    Sperber, D

    D. Sperber, D. Wilson, Relevance: Communication and Cognition, Blackwell, Oxford, 1986/1995

    1986

  32. [40]

    Sperber, D

    D. Sperber, D. Wilson, Representation and relevance, Mental (1988)

    1988

  33. [41]

    Steen, The paradox of metaphor: Why we need a three-dimensional model of metaphor, Metaphor and Symbol 23 (2008) 213–241

    G. Steen, The paradox of metaphor: Why we need a three-dimensional model of metaphor, Metaphor and Symbol 23 (2008) 213–241. doi:10.1080/10926480802426753

    2008 doi

  34. [42]

    G. J. Steen, Thinking by metaphor, fast and slow: Deliberate metaphor theory offers a new model for metaphor and its comprehension, Frontiers in Psychology Volume 14 - 2023 (2023). URL: https://www.frontiersin.org/journals/psychology/ articles/10.3389/fpsyg.2023.1242888. doi:1...

    2023

  35. [43]

    Michelli, X

    G. Michelli, X. Tong, E. Shutova, A framework for annotating and modelling intentions behind metaphor use, arXiv preprint arXiv:2407.03952 (2024)

    2024 arXiv

  36. [44]

    Gibbs, Intentions in the Experience of Meaning, Cambridge University Press, 1999

    R. Gibbs, Intentions in the Experience of Meaning, Cambridge University Press, 1999. URL: https://books.google.it/ books?id=FD1rVU4LtBgC

    1999

  37. [45]

    Austin, How to Do Things with Words, A galaxy book, Harvard University Press, 1962

    J. Austin, How to Do Things with Words, A galaxy book, Harvard University Press, 1962

    1962

  38. [46]

    J. L. Austin, A plea for excuses, in: J. L. Austin (Ed.), Philosophical Papers, 1961

    1961

  39. [47]

    Falkenhainer, K

    B. Falkenhainer, K. D. Forbus, D. Gentner, The structure-mapping engine: Algorithm and examples, Artificial Intelligence 41 (1989) 1–63. doi:10.1016/0004-3702(89)90077-5

    1989 doi

  40. [48]

    Speer, J

    R. Speer, J. Chin, C. Havasi, ConceptNet 5.5: An Open Multilingual Graph of General Knowledge, Proceedings of the AAAI Conference on Artificial Intelligence 31 (2017). URL: https://ojs.aaai.org/index.php/AAAI/article/view/11164. doi:10.1609/aaai.v31i1.11164

    2017 doi

  41. [49]

    S. C. Levinson, Pragmatics, Cambridge University Press, 1983

    1983

  42. [50]

    Popa-Wyatt, Compound figures: Priority and speech-act structure, Philosophical Studies 174 (2017) 141–161

    M. Popa-Wyatt, Compound figures: Priority and speech-act structure, Philosophical Studies 174 (2017) 141–161. doi:10.1007/s11098-016-0629-z

    2017 doi

  43. [51]

    De Giorgis, A

    S. De Giorgis, A. Gangemi, Efo: the emotion frame ontology, arXiv preprint arXiv:2401.10751 (2024)

    2024 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.