Temporal word embeddings in the study of metaphor change over time and across genres: a proof-of-concept study on English Veronica Mangiaterra1,* , Chiara Barattieri di San Pietro1 and Valentina Bambini1 1 Laboratory of Neurolinguistics and Experimental Pragmatics (NEPLab), Department of Humanities and Life Sciences, University School for Advanced Studies IUSS, Pavia, Italy Abstract Temporal word embeddings have been successfully employed in semantic change research to identify and trace shifts in the meaning of words. In a previous work, we developed an approach to study the diachrony of complex expressions, namely literary metaphors. Capitalizing on the evidence that measures of semantic similarity between the two terms of a metaphor approximate human judgments of the difficulty of the expression, we used time-locked measures of similarity to reconstruct the evolution of processing costs of literary metaphors over the past two centuries. In this work, we extend this approach previously used on Italian literary metaphors and we present a proof-of-concept study testing its crosslinguistic applicability on a set of 19th-century English literary metaphors. Our results show that the processing costs of metaphors changed as a function of textual genre but not of epoch: cosine similarity between the two terms of literary metaphors is higher in literary compared to nonliterary texts, and this difference is stable across epochs. Furthermore, we show that, depending on the metaphor structure, the difference between genres is affected by word-level variables, such as the frequency of the metaphor’s vehicle and the stability of the meaning of both topic and vehicle. In a broader perspective, general considerations can be drawn about the history of literary and nonliterary English language and the semantic change of words. Keywords metaphors, distributional semantics, temporal embeddings 1. Introduction veal a change in meaning [9]. Operatively, this means that by training vector space models on historical text Does the metaphor “The wind is a wrestler” convey the corpora from different epochs, it is possible to create same feeling today, as it did in the 1888 when Gerard time-locked representations of words: if the meaning of Manley Hopkins used it in the poem “That nature is a a word changed over time, its vectorial representation Heraclitean Fire and of the comfort of the Resurrection” at 𝑡1 will be different from its vectorial representation at [1]? The answer to this question is not trivial: human time 𝑡2 ; conversely, if the two vectors of the same word languages evolve constantly, alongside with the society at 𝑡1 and 𝑡2 are in close proximity, the meaning of the in which they are used, so much so that the concepts word has remained stable. Comparing words vectors di- associated with each word, as well as their semantic as- achronically, however, is not effortless and requires the sociations with other words, have changed to different temporal vector space models to be aligned. Alignment is degrees [2]. a crucial step in diachronic distributional semantics and it has been tackled by different approaches [10, 11, 12]. Studies on lexical semantic change have a long tra- Previous studies employing temporal embeddings have dition [3, 4] but, with the increasing availability of his- found that more frequent words change slower than torical language data and the development of new dig- less frequent words, and that polysemous words change ital tools, they radically opened up to new approaches faster than monosemous words [2], while synonyms tend coming from computational linguistics and distributional to change meaning comparably [13]. However, tempo- semantics [5, 6, 7]. In the diachronic declination of the ral word embeddings have been mostly applied to the Distributional Hypothesis [8], it is said that changes in study of the semantic change of single words and only the contexts in which a word occurs over time may re- marginally to complex linguistic expressions leaving the field with a knowledge gap on the evolution of meaning CLiC-it 2024: Tenth Italian Conference on Computational Linguistics, Dec 04 — 06, 2024, Pisa, Italy of a widespread linguistic and textual phenomenon such * Corresponding author. as, for instance, metaphors. $ veronica.mangiaterra@iusspavia.it (V. Mangiaterra); chiara.barattieridisanpietro@iusspavia.it (C. B. d. S. Pietro); Within the theoretical framework of Relevance Theory valentina.bambini@iusspavia.it (V. Bambini) [14], metaphors are non-literal uses of language involv-  0009-0001-7852-5259 (V. Mangiaterra); 0000-0003-4407-7037 ing a conceptual adjustment described as context-driven (C. B. d. S. Pietro); 0000-0001-5770-228X (V. Bambini) © 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License broadening of lexically denoted meaning of words. In Attribution 4.0 International (CC BY 4.0). CEUR ceur-ws.org Workshop ISSN 1613-0073 Proceedings terms of linguistic structure, metaphors normally involve its difficulty and we analyze how it varies across time two terms, the topic and the vehicle: for example, in and textual genres. We also consider the role of word fre- the metaphor ‘Sally is a chameleon’, the topic Sally is quency (WF) and vector coherence (VC), two widely used described by the broadened vehicle chameleon, to indi- measures in the study of semantic change [29, 30], as well cate a person who changes attitude/behavior to fit their as semantic neighborhood density (SND) in shaping the surroundings. While metaphors are broadly used in ev- difficulty of the expression. WF and VC were considered eryday communication, they are certainly a distinctive to assess the effect of the semantic change of the single feature of literary texts, as long evidenced in stylistics word on the evolution of whole metaphor understand- [15]. Past studies on literary metaphors, however, report ing, while SND was considered to analyze the impact of mixed results. The rating study by Katz et al. [16] found a measure known to synchronically impacts metaphor no difference between literary and everyday metaphors, understanding [31, 24] on its diachronic unfolding. while other studies showed that the former type is less familiar and more open-ended than the latter [17], but literary metaphors are rated as less difficult and more 2. Methods familiar when presented together with their original con- text [18]. Moreover, the processing of literay metaphors 2.1. Dataset of metaphors seems to be particularly effortful, given the multitude The study focuses on “classic” literary metaphors (i.e., of possible meanings they evoke [19]. Therefore, open metaphors found in 19th-century literary texts). In terms questions remain regarding how literary metaphors are of metaphor structure, we focused on metaphors in the processed. It must be also underlined that the literary form of ‘A is B’ (e.g. “Stars are dancers”) and ‘A of B’ metaphors used in previous studies were written tens or (e.g., “Clouds of melancholy”), as they clearly display hundreds of years ago. Yet, the effect of this diachronic di- the two metaphorical elements (topic and vehicle) and mension on their processing costs, as well as its interplayallow to avoid possible confounding factors (length of with textual genre in which metaphors are embedded, expression, intervening words, etc.). Twenty- four (24) remains an open question. ‘A is B’ metaphors were taken from the dataset in Katz et al. [16] and 115 metaphors in the form ‘A of B’ were In addition to its diachronic application, the use of retrieved from a collection of literary texts of the 19th vector space models can help characterize metaphors century. These latter were identified by PoS-tagging a thanks to the ability of these models to approximate hu- corpus of literary texts from the 19th century (see below) man performance in psycholinguistic tasks. Measures with spaCy [32], and then extracting only the ‘NOUN of derived from vector space models were shown to be NOUN’ constructions. The resulting list was then fur- able to approximate how humans process word meaning ther reduced by manually searching for words belonging [20, 21, 22] and, more specifically to correlate with how to known sources of metaphors, such as atmospheric humans perceive metaphorical expressions in terms of events (e.g., ‘rain’) or physical locations (e.g., ‘river’) [33], metaphoricity, difficulty, and other psycholinguistic di- following the methodology in Bambini et al. (2014) [18]. mensions [23, 24, 25]. In particular, semantic similarity, operationalized in vector space models as cosine sim- ilarity (CS) between topic and vehicle, has long been 2.2. Corpora and training considered relevant for metaphor studies [26] and, more To test whether the processing costs of metaphors recently, for automatic metaphor identification [27]. changed as a function of epoch, we collected corpora from the 19th century and from the 21st century. We In a previous study on Italian [28], we developed a also included different textual genres (literary vs. nonlit- novel method, employing the Temporal Word Embed- erary) of the corpora, to examine whether the difficulty dings with a Compass (TWEC) model [10] as training of the figurative expression is modulated by the stylis- procedure, to capture the temporal dynamics of literary tic features of different types of language. Following metaphors. This method combines the computational previous work [34], the corpora were built so as to be models’ abilities to approximate human judgments and representative of the language to which speakers of the their diachronic applications, allowing to track the di- two epochs were exposed, and specifically by combining achronic evolution of how literary metaphors are per- literary, nonfiction, and journalistic language for the 19th ceived by readers over the course of 200 years. In the century, and literary and web language (which includes present proof-of-concept study, we apply this approach sections of newspapers, blogs, and other text types that to English, to test its crosslinguistic applicability and can be found on the Internet) for the 21st century. Specif- whether it can provide language-specific insights into ically, we trained four diachronic vector space models on the evolution of metaphors. We take the similarity be- four corpora: tween the topic and vehicle of a metaphor as a proxy for • 19th-century literary corpus (32M tokens), con- sparsely distributed have a lower density. It is sisting of a collection of literary texts (both nar- computed as the mean cosine similarity between ratives and poetry) retrieved from the Gutenberg the target word and its 500 closest neighbors (stan- project (gutenberg.org); dard size from previous work, see [38]). • a 19th-century nonliterary corpus (25M tokens), • Vector coherence (VC): a measure of the stability consisting of nonliterary texts, such as magazines of a word’s meaning, computed as the cosine sim- or scientific essays, from the same online resource ilarity between the target word at 𝑡1 the target (gutenberg.org) word at 𝑡2 . Words with a high vector coherence • a 21st-century literary corpus (16M tokens), col- are considered to have stable meaning through lected from literary texts available on the web, time, while a low vector coherence means that employed without violating the “fair use” princi- the word’s meaning has changed. ple of copyright law; • Word frequency (WF): computed as the logarithm • a 21st-century nonliterary corpus (46M tokens), of the frequency of the target word in the refer- collected from portions of the UMBC web- Base ence corpus. corpus [35]. Each measure was collected for all the temporal slices, To train aligned temporal vector space models, we fol- extracted from the temporal vector space models (CS, lowed the procedure by Di Carlo et al. [10]. The TWEC SND, and VC) or corpora (WF). To analyze how the un- model is implemented on top of a Continuous Bag of derstanding of metaphors changed over time and if it was Words (CBOW) architecture [36]. The TWEC model ex- affected by genre and word-level variables, we fitted a ploits the double representation learned by the CBOW set of Linear Mixed Models (LMMs) using the R package model: the target matrix and the context matrix. First, a lme4 [39]. The two metaphorical structures were treated model, the so-called “compass”, is trained on the whole separately, fitting distinct models for ‘A is B’ and ‘A of B’ corpus, creating time-independent word embeddings. metaphors. The context matrix of the compass is then maintained fixed to train on each corpus a time- and genre-specific The linear mixed model considers CS as dependent target matrix from which we derive the temporal word variable and the interaction between epoch and genre embeddings. The four sets of embeddings obtained for and word-level variables as predictors. In all models the four corpora will represent the meaning of words Items (metaphors) were added as random variables. The in each time slice for the two genres. To validate our resulting formula was: models, following previous studies [2], we computed the lmer(cosine ∼ epoch * genre * (VC-topic + VC-vehicle + SND-topic + SND-vehicle + synchronic (within time period) accuracy of each vec- WF-topic + WF-vehicle) + (1|Item). tor space model against the MEN dataset [37], which Alpha level was set at .05. contains 3,000 pairs of words together with a semantic similarity score provided by humans. Finally, we tested 3. Results whether our measure of metaphor difficulty (cosine sim- ilarity between topic and vehicle) correlated with the First, to test the validity of the meaning representation in measure of difficulty in Katz et al. [16] dataset. the vector space models, we correlated the human scores of relatedness and the semantic similarity derived from 2.3. Measures of interest and analyses our word embedding for each pair of words in the MEN dataset [37] (Table 1). These results show strong corre- For each metaphor, we collected four measures of interest, lations, comparable to the results obtained by Hamilton at the metaphor- and word-level. et al. (2016) [2], indicating that the models accurately • Cosine similarity (CS): the similarity between the mimic humans’ representation of meaning (i.e., they have two terms of the metaphor (topic and vehicle). It a good synchronic accuracy). is computed as the cosine of the angle between 19th Literary 19th Nonliterary 21st Literary 21st Nonliterary the vectorial representations of the two words. .55 .58 .61 .59 CS is here considered as a proxy value of difficulty Table 1 of the metaphors. Results of correlation between models’ semantic similarity • Semantic neighborhood density (SND): a mea- scores and MEN dataset’s semantic similarity scores. All the sure of the density of the semantic space around correlation have a p < .001. a word. Words with many closely related words have a higher semantic density while Secondly, we tested whether cosine similarity between words whose neighbors are more distant and are the two terms of a metaphor correlated with the measure Figure 1: Effects of epoch and genre in defining the cosine similarity between the topic and vehicle of ‘A of B’ metaphors of difficulty from the dataset by Katz et al. [16]. Re- sults showed a moderate correlation (r(26) = .49, p < .05): metaphors with higher semantic similarity between topic and vehicle were rated with lower values of difficulty by participants, coherently with previous studies. Thirdly, we explored whether the change in the se- mantic similarity between the topics and the vehicles of literary metaphors is driven by the interaction be- tween the Epoch, Genre and single-word variables. The results of our predictors of interest are reported below. Concerning the ‘A of B’ metaphors’ mixed model, re- sults showed a main effect of genre (𝛽 = 0.81, t = 2.44, p = .01) and a significant three-way interaction between epoch, genre and vector coherence, both of the topic (𝛽 = 0.34, t = 2.018, p = .04) and of the vehicle (𝛽 = -1.715, t Figure 2: Effects of topic and vehicle VC in defining the cosine = -4.954, p < .001). These results indicate that the cosine similarity between the topic and vehicle of ‘A of B’ metaphors similarity of literary metaphors’ terms did not change over time, but it changed as a function of textual genres, resulting in greater difficulty (lower cosine similarity) in the vehicle (𝛽 = 0.06, t = 2.077, p = .04), but no main effects. nonliterary texts than in literary (Figure 1). As shown by The effect of WF of the vehicle showed different patterns the three-way interaction between Epoch and Genre and in the two time points and in the two genres (Figure the single-word variables in Figure 2, the effect of VC 3): while WF of the vehicle did not affect CS in literary acted differently in the two time points and in the two texts both in the past and in the present, more frequent genres. VC of the vehicle did not affect CS in literary and vehicles significantly increased CS in past nonliterary non- literary texts in the past; conversely, more stable texts and lowered CS in present nonliterary texts. vehicles significantly lowered CS in present literary texts and in- creased CS in present nonliterary texts. A similar trend can be observed for VC of the topic, where its stabil- 4. Discussion ity did not affect CS in the past, regardless of the literary genres. Conversely, stability of the topic contributed to In this proof-of-concept study, we characterized the tem- significantly increase CS in present literary texts, but less poral dynamics of a set of English literary metaphors to so in nonliterary texts. understand whether their processing costs changed over For ‘A is B’, the model revealed a significant three-way time. We also explored if this change was affected by the interaction between epoch, genre, and the frequency of genre of the texts, as well as by the semantic properties when found in literary contexts, compared to when en- countered in nonliterary texts. Hence, the difficulty of these metaphors is sensitive to the style of the text in which metaphors are found: when read in a text that has a literary style and aesthetic intent, the metaphor is less striking than the same metaphor in a nonliterary text. Moreover, we found a strong effect of the stability of the meaning of the vehicle in interaction with epoch and genre. This suggests that ‘A of B’ metaphors with more unstable vehicles are perceived as less difficult than ‘A of B’ metaphors with vehicles whose meanings remained stable over time. We interpreted this result in light of Traugott’s [41] theory of metaphorization, according to which the metaphorical use of a word can become one of Figure 3: Effects of vehicle WF in defining the cosine similar- its stable meanings. In the context of the present study, ity between the topic and vehicle of ‘A is B’ metaphors words that changed the most could have done so by incor- porating meanings derived from their metaphorical uses. As a result, when these unstable and broadened vehicles of the constituting elements of the metaphors (topic and are used, metaphors appear less difficult. The reader does vehicle). By leveraging on the diachronic applications of not need to broaden the concept expressed by the vehi- distributional semantics and extending a method already cle to interpret the metaphor, because the metaphorical applied to the study of Italian literary metaphors [28], nuances have entered the standard meaning of the word. we created a series of time-locked semantic representa- From a qualitative observation of the data, we can notice, tions of 139 English metaphors, from which we derived for instance, that a metaphor such as “Wave of horror”, a measure of the cosine similarity between their terms where the vehicle wave incorporated the meaning of ‘sud- (CS), taken as a proxy of their difficulty, together with den increase in a particular phenomenon’, is perceived as semantic neighborhood density (SND), stability over time less metaphorical than “Clouds of doubt”, whose vehicle (VC), and, from four diachronic corpora, frequency (WF) clouds has maintained its original meaning. of their topics and vehicles. For ‘A is B’ metaphors, instead, the statistical model Results showed no effect of epoch for either ‘A is highlighted an effect of the frequency of the vehicle in B’ or ‘A of B’ literary metaphors. Thus, no noticeable interaction with epoch and genre. In nonliterary texts, change in CS over time was revealed, suggesting that the perceived difficulty of ‘A is B’ metaphors differed as these metaphors come with similar processing costs for a function of the WF of their vehicle, to the point that contemporary readers and for readers of the epoch in metaphors showed opposite patterns in the past and in which the metaphors were created. The absence of an the present: in the past, the less frequent the vehicle, effect of epoch can be better understood by consider- the more metaphorical the whole metaphorical expres- ing the historical evolution of the English language, and sion; in the present, the less frequent the vehicle, the less specifically its early standardization. As stated by Wyld metaphorical the metaphor. The pattern found in the [40], literary writing as early as the 18th century was 19th-century space model is in line with previous studies considered ‘English of our own age in all its essentials’. [42] that found that metaphors with less frequent vehicles In line with this consideration, our results point to the are regarded as more metaphorical than those with highly stability of the main stylistic features of the English lan- frequent vehicles, indicating that the most metaphorical guage in the last two centuries, including those related metaphors are those in which the vehicle communicates to metaphors. something new about the topic. Going back to Hopkins’ metaphor "The wind is a wrestler", the vehicle wrestler, While literary metaphors are not processed differently as a particularly low frequency word in the 19th century, based on the epoch, the influence of textual genre is was indeed communicating something new about the noticeable. This factor emerged both as a main effect topic wind. As such, the metaphors might have been per- and in different interaction patterns with single-word ceived as more difficult and “more metaphorical”, leading variables, varying according to the type of metaphor. to the creation of a new concept. The very same metaphor is nowadays perceived differently, because the frequency For ‘A of B’ metaphors, results revealed that the dif- of the vehicle has changed: wrestler has become more ficulty of these metaphors changed as a function of the frequent, and the whole expression has lost some of its genre. In particular, they are perceived as less difficult metaphoricity for the 21st-century readers. Overall, our results suggest that for the English lan- 7. Acknowledgment guage, metaphor processing costs are not affected by the temporal distance between the creation of metaphors, This work received support from the European Re- which occurred in the 19th century, and their processing search Council under the EU’s Horizon Europe program, by today’s readers. Instead, the key factor modulating ERC Consolidator Grant “PROcessing MEtaphors: Neu- metaphor processing costs seems to be the textual genre rochronometry, Acquisition and Decay, PROMENADE” in which they appear. This modulation, however, occur [101045733]. The content of this article is the sole respon- to a different extent depending on the syntactic structure sibility of the authors. The European Commission or its of the metaphors and in interaction with single word mea- services cannot be held responsible for any use that may sures. Indeed, we observe that in defining what drives be made of the information it contains. the difficulty of metaphors, different patterns emerged for the ‘A of B’ and ‘A is B’ structures. While for the for- mer, in addition to the main effect of genre, we found the References effect of vector coherence in interaction with epoch and [1] W. H. Gardner, N. H. MacKenzie (Eds.), The Poems genre, for the latter the diachronic evolution of metaphor of Gerard Manley Hopkins, Oxford University Press, processing costs is related to the interaction of word fre- 1967. quency with epoch and genre. [2] W. L. Hamilton, J. Leskovec, D. Jurafsky, Diachronic While these differences might reflect genuine effects of word embeddings reveal statistical laws of semantic the syntactic structure and how it impacts metaphorical change, arXiv preprint arXiv:1605.09096 (2016). predication [43, 44, 45], we must acknowledge that the [3] E. C. Traugott, R. B. Dasher, Regularity in semantic numerosity of the two sets of items varies and this might change, Cambridge University Press, 2002. obscure some of the effects in the less represented type [4] B. W. Fortson IV, An approach to semantic change, (A is B). Future studies are needed to further explore the in: B. D. Joseph, R. D. Janda (Eds.), The Hand- whole range of diachronic changes in processing related book of Historical Linguistic., Blackwell Publishing, to structural differences. Malden, MA, 2003, pp. 648–666. [5] A. Kutuzov, L. Øvrelid, T. Szymanski, E. Velldal, In conclusion, this proof-of-concept study proposed an Diachronic word embeddings and semantic shifts: adaptation from Italian to English of a method employ- a survey, arXiv preprint arXiv:1806.03537 (2018). ing temporal word embeddings to study the evolution of [6] N. Tahmasebi, L. Borin, A. Jatowt, Survey of com- metaphors. Thanks to this approach, we could elucidate putational approaches to lexical semantic change, that the processing costs of English literary metaphors arXiv preprint arXiv:1811.06278 (2018). is stable over time (differently from Italian) but is dy- [7] X. Tang, A state-of-the-art of semantic change com- namically affected by stylistic features of texts and by putation, Natural Language Engineering 24 (2018) single-word measures. The proposed method seems to 649–676. doi:10.1017/S1351324918000220. be sensitive to the specificities of the language under [8] Z. Harris, Distributional hypothesis, Word World investigation, supporting its crosslinguistic applicability. 10 (1954) 146–162. [9] M. Hilpert, Germanic future constructions: A usage- 5. Ethic statement based approach to language change, John Ben- jamins Publishing, Amsterdam, 2008. The work aims to use computational tools for the study [10] V. Di Carlo, F. Bianchi, M. Palmonari, Training of literature, thus enhancing the literary heritage with temporal word embeddings with a compass, in: innovative methods that can provide insights for scholars Proceedings of the AAAI conference on artificial from a wide range of disciplines. We are aware, however, intelligence, volume 33, 2019, pp. 6326–6334. that the corpora used are not representative of the entire [11] K. Gulordava, M. Baroni, A distributional similarity spectrum of varieties of English, but of educated, Western approach to the detection of semantic change in English. Hence, our results may not coincide with the the google books ngram corpus., in: Proceedings of general evolution of the language but provide a partial the GEMS 2011 workshop on geometrical models view of it. of natural language semantics, 2011, pp. 67–71. [12] V. Kulkarni, R. Al-Rfou, B. Perozzi, S. Skiena, Statis- tically significant detection of linguistic change, in: 6. Data availability Proceedings of the 24th international conference on world wide web, 2015, pp. 625–635. Temporal vector space models and metaphor datasets [13] Y. Xu, C. Kemp, A computational evaluation of used in the study are available at https://osf.io/j8bd7/ two laws of semantic change., in: Proceedings of ?view_only=4cd623d5622b4ed0bd1624c42aff0f40$. the 37th Annual Meeting of the Cognitive Science synesthetic metaphors, Metaphor and the Social Society, 2015, p. 2703–2708. World 13 (2023) 59–80. doi:10.1075/msw.00029. [14] D. Wilson, R. Carston, A unitary approach to lex- win. ical pragmatics: Relevance, inference and ad hoc [26] A. N. Katz, A. Paivio, M. Marschark, Poetic com- concepts, in: N. Burton-Roberts (Ed.), Pragmatics, parisons: Psychological dimensions of metaphoric Palgrave Macmillan, Basingstoke, 2007, p. 230–259. processing, Journal of Psycholinguistic Research [15] M. Fludernik, D. C. Freeman, M. H. Freeman, 14 (1985) 365–383. doi:https://doi.org/10. Metaphor and beyond: An introduction, Poetics 1007/BF01067881. today 20 (1999) 383–396. [27] E. Shutova, Design and evaluation of metaphor [16] A. N. Katz, A. Paivio, M. Marschark, J. M. processing systems, Computational Linguis- Clark, Norms for 204 literary and 260 nonliter- tics 41 (2015) 579–623. doi:https://doi.org/10. ary metaphors on 10 psychological dimensions, 1162/COLI_a_00233. Metaphor and Symbol 3 (1988) 191–214. doi:10. [28] V. Mangiaterra, C. Barattieri di San Pietro, V. Bam- 1207/s15327868ms0304_1. bini, How literary metaphors change over two [17] E. Semino, G. Steen, Metaphor in literature, in: centuries (in prep). J. R. W. Gibbs (Ed.), The Cambridge handbook of [29] A. Englhardt, J. Willkomm, M. Schäler, K. Böhm, metaphor and thought, Cambridge University Press, Improving semantic change analysis by combining Cambridge, 2008, pp. 232–246. word embeddings and word frequencies, Interna- [18] V. Bambini, D. Resta, M. Grimaldi, A dataset of tional Journal on Digital Libraries 21 (2020) 247–264. metaphors from the italian literature: Exploring doi:10.1007/s00799-019-00271-6. psycholinguistic variables and the role of context, [30] Q. Feltgen, B. Fagard, J.-P. Nadal, Frequency pat- PloS one 9 (2014) e105634. doi:10.1371/journal. terns of semantic change: corpus-based evidence of pone.0105634. a near-critical dynamics in language change, Royal [19] V. Bambini, P. Canal, D. Resta, M. Grimaldi, Time Society open science 4 (2017) 170830. doi:10.1098/ course and neurophysiological underpinnings of rsos.170830. metaphor in literary context, Discourse Processes [31] H. Al-Azary, A. N. Katz, On choosing the vehicles 56 (2019) 77–97. doi:10.1080/0163853X.2017. of metaphors 2.0: the interactive effects of seman- 1401876. tic neighborhood density and body-object interac- [20] S. Bhatia, R. Richie, W. Zou, Distributed seman- tion on metaphor production, Frontiers in Psychol- tic representations for modeling human judgment, ogy 14 (2023) 1216561. doi:10.3389/fpsyg.2023. Current Opinion in Behavioral Sciences 29 (2019) 1216561. 31–36. doi:10.1016/j.cobeha.2019.01.020. [32] M. Honnibal, I. Montani, spacy 2: Natural [21] F. Günther, C. Dudschig, B. Kaup, Latent semantic language understanding with bloom embeddings, analysis cosines as a cognitive similarity measure: convolutional neural networks and incremental Evidence from priming studies, Quarterly Jour- parsing (2017). doi:https://doi.org/10.5281/ nal of Experimental Psychology 69 (2016) 626–653. zenodo.3358113. doi:10.1177/174569161986137. [33] P. Hanks, Metaphoricity is gradable, Trends [22] M. N. Jones, J. Willits, S. Dennis, M. Jones, Models in Linguistics Studies and Monographs 171 of semantic memory, in: J. Busemeyer, Z. Wang, (2006) 17. doi:https://doi.org/10.1515/ J. Townsend, A. Eidels (Eds.), Oxford handbook of 9783110199895.17. mathematical and computational psychology, Ox- [34] V. Bambini, M. Trevisan, Esploracolfis: ford University Press, New York, NY, 2015, pp. 232– Un’interfaccia web per le ricerche sul cor- 254. pus e lessico di frequenza dell’italiano scritto [23] S. McGregor, K. Agres, K. Rataj, M. Purver, (colfis), Quaderni del Laboratorio di Linguistica 11 G. Wiggins, Re-representing metaphor: Modeling (2012) 1–16. metaphor perception using dynamically contextual [35] L. Han, A. L. Kashyap, T. Finin, J. Mayfield, J. Weese, distributional semantics, Frontiers in psychology Umbc_ebiquity-core: Semantic textual similarity 10 (2019) 765. doi:10.3389/fpsyg.2019.00765. systems, in: Second Joint Conference on Lexical [24] N. J. Reid, H. Al-Azary, A. N. Katz, Cognitive fac- and Computational Semantics (* SEM), Volume 1: tors related to metaphor goodness in poetic and Proceedings of the Main Conference and the Shared non-literary metaphor, Metaphor and Symbol Task: Semantic Textual Similarity, 2013, pp. 44–52. 38 (2023) 130–148. doi:10.1080/10926488.2021. [36] T. Mikolov, K. Chen, G. Corrado, J. Dean, Efficient 2011285. estimation of word representations in vector space, [25] B. Winter, F. Strik-Lievers, Semantic distance pre- arXiv preprint arXiv:1301.3781 (2013). dicts metaphoricity and creativity judgments in [37] E. Bruni, G. Boleda, M. Baroni, N.-K. Tran, Distri- butional semantics in technicolor, in: Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), 2012, pp. 136–145. [38] W. Kintsch, Metaphor comprehension: A compu- tational theory, Psychonomic bulletin & review 7 (2000) 257–266. doi:10.3758/BF03212981. [39] D. Bates, M. Mächler, B. Bolker, S. Walker, Fit- ting linear mixed-effects models using lme4, arXiv preprint arXiv:1406.5823 (2014). [40] H. C. Wyld, A history of modern colloquial English, Basil Blackwell, 1936. [41] E. C. Traugott, Semantic change, in: Ox- ford research encyclopedia of linguistics, Oxford University Press, 2017. doi:10.1093/acrefore/ 9780199384655.013.323. [42] J. Littlemore, P. Pérez Sobrino, D. Houghton, J. Shi, B. Winter, What makes a good metaphor? a cross- cultural study of computer-generated metaphor ap- preciation, Metaphor and Symbol 33 (2018) 101–122. doi:10.1080/10926488.2018.1434944. [43] V. Bambini, M. Ghio, A. Moro, P. B. Schumacher, Differentiating among pragmatic uses of words through timed sensicality judgments, Frontiers in psychology 4 (2013) 938. doi:https://doi.org/ 10.3389/fpsyg.2013.00938. [44] R. Carston, X. Yan, Metaphor processing: Referring and predicating, Cognition 238 (2023) 105534. doi:https://doi.org/10.1016/ j.cognition.2023.105534. [45] E. Tonini, L. Bischetti, P. Del Sette, E. Tosi, S. Lecce, V. Bambini, The relationship between metaphor skills and theory of mind in middle childhood: Task and developmental effects, Cognition 238 (2023) 105504. doi:https://doi.org/10.1016/ j.cognition.2023.105504.