From critical analysis to formal representation: literary characters, interpretations, and ontologies
Dipartimento di lettere e filosofia, via Tommaso Gar, 14, 38122 Trento TN, 21-23 October 2025
Online as well as onsite participation will be possible. Please let us known through the following form.
For information relative to the workshop, please write to: miteprinpnrr@gmail.com
Tentative program (to be refined; see abstracts below )
DAY 1 (21/10)
(discussants: Roberta Ferrario and Iuris Mocchiutti)
9:30-10:15 Introduction and presentation of MITE
10:15-11:00 Evelyn Gius (Technical University of Darmstadt) Computational Narratology between Text Surface, Deeper Structure and Readers
11:00-11:30 BREAK
(discussant: Alessandro Mosca)
11:30 – 12:15 Federico Pianzola (University of Groningen) The GOLEM Ontology for Narrative and Fiction
12:15-13:00 Sara Tonelli (Fondazione Bruno Kessler) Human Label Variation in Linguistic Annotation
13:00-14:30 Lunch BREAK
(discussant: Gaia Tomazzoli)
14:30-15:15 Stefano Ballerio (University of Milan) «He was one of us». Sul giudizio morale in Lord Jim di Joseph Conrad
15:15-16:00 Heloísa Abreu de Lima (Sapienza University of Rome) Can we trust Fiammetta? Moral judgment and narrative reliability in Boccaccio’s “Elegia di Madonna Fiammetta”
16:00-16:30 BREAK
16:30-17:15 Emanuele Bottazzi (CNR ISTC), Interpreting is not understanding: how machines get lost in the forest of symbols
17:15-18:00 Panel (tba) (discussant: Claudio Masolo)
DAY 2 (22/10)
(discussant: Heloísa Abreu de Lima)
9:30-10:15 Elena Porciani (Università della Campania ‘Luigi Vanvitelli’) L’invenzione delle personagge. Ragioni teoriche e pratica genealogica
10:15-11:00 Gaia Tomazzoli (Sapienza University of Rome) Authorial stance in literary criticism: a cross-linguistic and cross-gender analysis
11:00-11:30 BREAK
(discussant: Jansan Favazzo )
11:30 – 12:15 Enrico Terrone (University of Genoa) The Experiential Function of Fictional Characters
12:15-13:00 Michele Paolini Paoletti (University of Macerata) Ficta and their Powerful Identity: A Case Study from Paul Auster’s 4 3 2 1
13:00-14:30 Lunch BREAK
(discussant: Marco Buzzoni)
14:30-15:15 Francesco Orilia (University of Macerata) tba
15:15-16:00 tba
16:00-16:30 BREAK
16:30-17:15 Jansan Favazzo (University of Macerata), Fiction, Identity, and Existence: A Study in Experimental Philosophy
17:15-18:00 Panel tba
DAY 3 (23/10)
(discussant: Claudio Masolo)
9:30-10:15 Carlo Meghini (CNR ISTI) tba
10:15-11:00 Emilio M. Sanfilippo (CNR ISTC) tba
11:00-11:15 BREAK
11:15 – 12:00 Final discussion (all)
Heloísa Abreu de Lima, Sapienza University of Rome
Title: Can we trust Fiammetta? Moral judgment and narrative reliability in Boccaccio’s “Elegia di Madonna Fiammetta”
Fiammetta, the enigmatic woman celebrated in various works by Giovanni Boccaccio, represents one of the most compelling yet contradictory female characters in Trecento literature. Her multifaceted portrayals have sparked critical debates that illuminate key aspects of Boccaccio’s oeuvre. In this talk, I intend to examine a long-standing critical debate regarding Fiammetta, analyzing its implications for the interpretation of Boccaccio’s Elegia di Madonna Fiammetta. First, I will trace Fiammetta’s various appearances across Boccaccio’s works, highlighting not only the evident inconsistencies between her portrayals, but also the elements that, on the other hand, point to an underlying continuity from one work to another. Then, I will focus specifically on the Elegia di Madonna Fiammetta, where Fiammetta appears both as protagonist and as first-person narrator. In particular, I will address the critical problems related to the moral judgments that the character may be subject to. As I argue, these moral judgments shape our interpretation of Fiammetta’s reliability as the narrator of her own story, thereby determining how we engage with the text’s meaning and the author’s potential stance toward the character.
Stefano Ballerio, Università degli Studi di Milano
Title: «He was one of us». Sul giudizio morale in Lord Jim di Joseph Conrad
La mia relazione verte su Lord Jim (1900) di Joseph Conrad e in particolare sul suo protagonista Jim. Vorrei soffermarmi sui diversi giudizi morali dei quali Jim è oggetto, sia nella finzione romanzesca, sia nella ricezione. Per il gesto che determina il suo destino, l’abbandono del Patna e dei suoi passeggeri, Jim è ripetutamente giudicato da numerosi personaggi che incontra nel seguito della sua storia o che di quella storia, anche senza incontrarlo, vengono a sapere, e inoltre dal suo narratore Marlow e da se stesso. Al di là della finzione romanzesca, gli interpreti dell’opera hanno discusso a loro volta la figura morale di Jim. Vorrei osservare questi diversi giudizi e riflettere su come funzioni il giudizio morale nella narrativa romanzesca. Oltre alla letteratura critica sul romanzo e sulla narrativa di Conrad in generale, i riferimenti principali potrebbero essere Michail Bachtin, con le sue considerazioni sul rapporto fra autore ed eroe, per la questione dei giudizi morali all’interno della finzione, e l’ermeneutica gadameriana, per la questione dei giudizi morali da parte dei lettori.
Emanuele Bottazzi (CNR ISTC)
Title: Interpreting is not understanding: how machines get lost in the forest of symbols
Current computational systems struggle to represent literary interpretive debates reliably, due to fundamental conceptual limitations. Despite impressive recent advances, we point out crucial obstacles that prevent us from simply feeding texts and interpretations into a system and expecting faithful models of interpretive discourse. A first obstacle lies in the relation between understanding and interpretation, which is complex and bidirectional. Sometimes interpretation presupposes understanding: one cannot interpret a novel, or criticism of that novel, without first achieving some degree of textual understanding. At other times, interpretation is the only resource precisely because understanding fails: when a passage seems ambiguous, claiming to have understood it forces us to choose among rival readings—yet often we cannot make that choice decisively. This recalls Wittgenstein’s rule-following problem. Following a rule is, at bottom, a matter of practice, not interpretation, since many interpretations are legitimate and thus leave us unsure how to proceed. This reveals why understanding differs fundamentally from interpreting: understanding is about being properly oriented within a text’s world, not about producing readings of it. This leads us to the second obstacle. We argue that understanding a text, like orienting oneself in space, is a fundamentally de se process: it concerns the way we ourselves are situated within the text or the space. In contrast, artificial systems maintain an invariably external perspective, preventing them from navigating the sometimes rigid, sometimes elastic constraints needed to comprehend a text and to sustain dialogue between texts.
Bottazzi Grifoni, E., & Ferrario, R. (2025). The bewitching AI: The Illusion of Communication with Large Language Models. Philosophy & Technology, 38(2), 61. https://doi.org/10.1007/s13347-025-00893-6
Jansan Favazzo (University of Macerata)
Title: Fiction, Identity, and Existence: A Study in Experimental Philosophy
The category of fictional entities (ficta, in short) is highly controversial and extensively debated in contemporary metaphysics. Within the MITE project, we conducted an empirical study through questionnaire administration aimed at testing the following hypotheses. With respect to the identity of ficta, (1a) the construct validity of our theoretical assessment, and (1b) whether it can be observed a commonsensical response to the identity conundrums raised by ficta. With respect to the existence of ficta, (2) whether it can be identified a commonsensical opinion about the ontological and metaphysical status of ficta (respectively, whether ficta are genuine entities and, if so, what kind of entities they are). Moreover, with respect to both, (3) whether there is a significant difference in response between laypeople and people with a strong background either in the philosophical or in the literary field.
Evelyn Gius (Technical University of Darmstadt)
Title: Computational Narratology between Text Surface, Deeper Structure and Readers
For analyzing plot-related phenomena, we have been working on different levels of granularity with relation to different features of the narrative in the past years. In this talk, I will present this work with a focus on overarching questions of operationalization. Our approach takes into account the narratological modeling of narratives which ranges from analysis of textual representation over the identification of deep structures to questions of reader responses. The idea is that this is also a more successful approach to automated analysis of narratives than approaches relying on textual analysis based on surface features only. We therefore combine textual features with general narrative („deep“) structure as well as, to some extent, reader’s perception.
Gerstorfer, Dominik, and Evelyn Gius. 2025. “Operationalizing Operationalizing.” In DHd2025, 345–48. Bielefeld. https://doi.org/10.5281/zenodo.14943080.
Gius, Evelyn, and Michael Vauth. 2022. “Towards an Event Based Plot Model. A Computational Narratology Approach.” Journal of Computational Literary Studies 1 (1). https://doi.org/10.48694/jcls.110.
Hatzel, Hans Ole, Haimo Stiemer, Chris Biemann, and Evelyn Gius. 2023. “Machine Learning in Computational Literary Studies.” Edited by Hubert Mara and Ricardo Usbeck. It – Information Technology, August. https://doi.org/10.1515/itit-2023-0041.
Michele Paolini Paoletti (University of Macerata)
Title: Ficta and their Powerful Identity: A Case Study from Paul Auster’s 4 3 2 1
In his novel 4 3 2 1, Paul Auster explores four possible lives of Archie Ferguson, a Jewish guy born in March 1947 and living in New Jersey. The novel aims at showing how luck and coincidences may shape one’s life, leading to radically different outcomes. But – to philosophers – it also presents a dilemma about the identity of fictional characters (or ficta) within one and the same work of fiction. Indeed, if Archie Ferguson is numerically the same fictum across all of his possible lives, he turns out to have radically different and inconsistent features. On the other hand, if each ‘version’ of Archie Ferguson actually is a fictum in its own right, this is literally at odds with there being multiple ‘versions’ of the same fictum. In this talk, I shall suggest that there is a way to preserve the numerical identity of Archie Ferguson as a fictum across all of his possible lives. This consists in allowing that the identity of at least some ficta is also or only fixed by the powers they have according to the relevant stories, i.e., by properties of the form: being able to do M/become M (in circumstances C). More precisely, the identity of Archie Ferguson across all of his possible lives is also or only fixed by certain powers that have certain features: first, they have relatively generic manifestations (thus allowing for the realization of distinct subtypes of manifestations falling under the same type); secondly, they are endowed with specific degrees of ‘strength’, grounding the higher/lesser probabilities that such powers get manifested; thirdly, such degrees of ‘strength’ (and the corresponding probabilities) may be affected by external circumstances.
Elena Porciani (Università della Campania ‘Luigi Vanvitelli’)
Title: L’invenzione delle personagge. Ragioni teoriche e pratica genealogica
Nel mio intervento proporrò una riflessione sul concetto di personaggia, ‘inventato’ dalla Società Italiana delle Letterate a partire dal 2009, adottato da alcune studiose in una prospettiva letteraria e visuale di genere, ma mai sottoposto a un’effettiva elaborazione teorica. Mi propongo, in particolare, di indagare le ragioni d’essere della personaggia individuandone la specificità in una dimensione genealogico-femminista che costituisce una variante di genere nei fenomeni di costruzione e ricezione dei tesi narrativi.
Federico Pianzola (University of Groningen)
Title: The GOLEM Ontology for Narrative and Fiction
This talk introduces the GOLEM ontology, a novel framework designed to provide a structured and computationally tractable representation of narrative and fictional elements. Addressing limitations in existing ontologies regarding the integration of fictional entities and diverse narrative theories, our model extends CIDOC CRM and LRMoo, and leverages DOLCE’s cognitive foundations to provide a flexible and interoperable framework. The ontology captures complexities of narrative structure, character dynamics, and fictional worlds, while supporting provenance tracking and pluralistic interpretations. The modular structure facilitates alignment with various literary and narrative theories and integration of external resources. Future work will focus on expanding domain-specific extensions, validating the model through larger-scale case studies, and developing a reader response module to systematically model the reception of narratives. By fostering interoperability between literary theory, fan cultures, and computational analysis, this ontology lays a foundation for interoperable comparative research on narrative and fiction.
Project website: https://golemlab.eu
SPARQL endpoint: http://graph.golemlab.eu:8890/sparql
Web interface for the knowledge graph: http://search.golemlab.eu:3006
Ontology for Narrative and Fiction: https://github.com/GOLEM-lab/golem-ontology/wiki
Enrico Terrone (University of Genoa)
Title: The Experiential Function of Fictional Characters
If one assumes both the ontological view according to which fictional characters are abstract artifacts and the metaphysical view according to which artifacts essentially have functions, one should conclude that fictional characters have functions and specify which those are. I will argue that fictional characters have the function of generating peculiar configurations of mental states that involve not only imagination and emotions directed toward the fictional character as an imaginary flesh-and-blood individual but also attention to the structural and genetic features of the fictional character as an artifact.
(2023) “Taking Abstract Artifacts Seriously. The Functioning and Malfunctioning of Fictional Characters”, in “Fiction and Metaphysics”, eds. Frederick Kroon and Alberto Voltolini, Philosophies, 2023, 8, 6, 105, DOI: https://doi.org/10.3390/philosophies8060105
(2021) “Twofileness. A Functionalist Approach to Fictional Characters and Mental Files”, Erkenntnis, 86, pp. 129-147, DOI: http://dx.doi.org/10.1007/s10670-018-0097-2
(2017) “On Fictional Characters as Types”, British Journal of Aesthetics, 57, 2, pp. 161-176.
Gaia Tomazzoli (Sapienza University of Rome)
Title: Authorial stance in literary criticism: a cross-linguistic and cross-gender analysis
This paper aims to investigate authorial stance in modern literary criticism. It relies on a small corpus of scholarly essays devoted to selected female characters in Medieval Italian literature: the corpus will include texts written over the past 150 years, evenly distributed across three different linguistic, cultural, and methodological landscapes – Italian, French, and English – and authored by both men and women. Through a manual annotation of this corpus, I will examine how gender and linguistic-cultural background influence the way critics express personal opinions, assert interpretive claims, and negotiate their authorial stance.
Specifically, my analysis draws on the stance system developed by Hyland, and thus considers strategies of hedging and boosting employed to qualify statements, as well as explicit self-reference vs the use of impersonal constructions. Through a comparative approach, the paper seeks to uncover systematic patterns that reflect broader cultural and gendered norms governing scholarly authority and the presentation of interpretation as either subjective or objective. By tracing these dimensions across texts written by both male and female authors coming from different linguistic, cultural, and disciplinary backgrounds. The ultimate goal is to contribute to a deeper understanding of how critical discourse constructs its claims to legitimacy and how implicit biases may shape the portrayal of Medieval female figures in literary criticism.
Sara Tonelli (Fondazione Bruno Kessler)
Title: Human Label Variation in Linguistic Annotation
Human label variation (Plank, 2022) reflects different aspects of linguistic annotation, from annotators’ subjectivity to sloppiness or lack of enough context to interpret a text, as well as genuine disagreement due to multiple plausible labels. However, this poses several challenges concerning both the interpretation of the annotated linguistic datasets and the implementation of annotation systems that should be trained on “gold data”. In this talk I will present ongoing research in the NLP community around the topic of human label variation, covering new practices for linguistic annotation, shared tasks, evaluation methodologies and classification approaches.
Alan Ramponi, Agnese Daffara, and Sara Tonelli. 2025. Fine-grained Fallacy Detection with Human Label Variation. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), pages 762–784, Albuquerque, New Mexico. Association for Computational Linguistics. https://aclanthology.org/2025.naacl-long.34/
Marta Sandri, Elisa Leonardelli, Sara Tonelli, and Elisabetta Jezek. 2023. Why Don’t You Do It Right? Analysing Annotators’ Disagreement in Subjective Tasks. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 2428–2441, Dubrovnik, Croatia. Association for Computational Linguistics. https://aclanthology.org/2023.eacl-main.178/
Elisa Leonardelli, Stefano Menini, Alessio Palmero Aprosio, Marco Guerini, and Sara Tonelli. 2021. Agreeing to Disagree: Annotating Offensive Language Datasets with Annotators’ Disagreement. In Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 10528–10539, Online and Punta Cana, Dominican Republic. Association for Computational Linguistics. https://aclanthology.org/2021.emnlp-main.822/