Pith. sign in

REVIEW 3 major objections 4 minor 1 cited by

Characterizing the Investigative Methods of Fictional Detectives with Large Language Models

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A five-phase LLM workflow derives consensus trait profiles for seven fictional detectives, and reverse identification succeeds in 91.43% of trials.

desk verdict A transparent multi-LLM workflow for detective trait profiles, but the 91.43% accuracy is ensemble self-consistency rather than external validation; worth a serious referee, not a desk reject. read the letter →

arxiv 2505.07601 v1 pith:3QN3RMBK submitted 2025-05-12 cs.CL cs.AI

Edirlei Soares de Lima , Marco A. Casanova , Bruno Feijó , Antonio L. Furtado This is my paper · ORCID
classification cs.CLcs.AI
keywords detectivefictioncomputationalnarratologylargelanguagemodelsinvestigativetraitsreverseidentificationcharacteranalysisnarrativegenerationLLMensemble
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

This paper claims that a multi-phase workflow built from fifteen large language models can systematically capture the distinctive investigative methods of fictional detectives and compress them into reusable trait profiles. The workflow was run on seven iconic detectives — Hercule Poirot, Sherlock Holmes, William Murdoch, Columbo, Father Brown, Miss Marple, and Auguste Dupin — and the resulting profiles were checked against existing literary analyses. The authors then reversed the process: each profile was shown to all fifteen models, which had to name the detective from the trait list alone, and 91.43% of those identifications were correct. If the claim holds, computational narratology gains a scalable way to turn character interpretation into structured data that story-generation systems can reuse.

What carries the argument

The load-bearing mechanism is a five-phase LLM ensemble workflow. Each of fifteen models independently writes a constrained five-sentence description of a detective's method; a single extraction model turns each description into a bullet list of traits; a grouping model merges semantically equivalent traits across models; a consensus threshold keeps only trait groups supported by at least 20 percent of the models (three of fifteen); and the surviving trait list is fed back to all fifteen models, which must identify the detective from the list alone. The reverse-identification phase is the step that makes distinctiveness measurable rather than merely asserted.

What would settle it

A concrete check is to run the reverse-identification test on a fresh set of detectives that the fifteen models did not help profile — for example, two detectives created from the workflow's own trait lists with labels withheld — and see whether accuracy stays near 91 percent; if it collapses, the original score was measuring memorized canon rather than trait distinctiveness. A sharper version would present the same trait lists with one trait removed at a time and measure how much each trait contributes to correct identification.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the multi-model consensus process yields trait lists specific enough for models to recover the detective's identity from the traits alone. Across 105 reverse-identification trials, 96 were correct. Four detectives — Poirot, Columbo, Father Brown, and Miss Marple — were identified perfectly by all fifteen models; Sherlock Holmes and William Murdoch each had one misclassification; Auguste Dupin was the weak point, identified correctly only 8 times out of 15, with most errors placing him as Sherlock Holmes, a confusion the paper attributes to Dupin's documented role as the prototype for Holmes. The authors read this pattern as evidence that the profiles capture real investigative styles, including their historical overlaps.

Load-bearing premise

The load-bearing premise is that an LLM correctly naming a detective from a trait list is evidence that the traits are distinctive and accurate, even though the same models generated those traits from their own training knowledge of the same seven detectives.

Editorial extensions

If this is right

  • The synthesized trait profiles are reusable character models: a narrative generation system could use them to keep a detective acting in character across a generated story.
  • The workflow is prompt-generic, so the same five phases can be pointed at other detectives, at non-detective roles such as criminals or victims, or at characters in other genres.
  • Because the consensus filter drops traits supported by only one model, the final profiles represent cross-model agreement rather than any single model's idiosyncratic reading.
  • The Dupin/Holmes confusion suggests that near-neighbour detective pairs will be the limiting case of the method, and that is where profile refinement would matter most.
  • The reported 91.43% accuracy is a property of this seven-detective set; adding more detectives would make the identification task harder and test the profiles' discriminative power more severely.

Reading between the lines

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

  • The reverse-identification test would be stronger if run against a control set of non-canonical or lesser-known detectives, because the seven chosen characters are heavily represented in LLM training data; a high score could then be partly recognition rather than distinctiveness.
  • Stability of the profiles across different LLM panels is testable: repeating the workflow with a different set of models would show whether the consensus traits drift, and drift would indicate which traits are model-dependent rather than character-dependent.
  • The semantic grouping step deliberately generalizes specific reasoning terms, so the final profiles can lose the vocabulary of reasoning type (for example, abduction disappears under 'logical reasoning'); a profile intended for narrative generation might want to retain such distinctions.
  • A direct extension would use these profiles as conditioning input in a story generator and measure whether readers can identify which detective is acting, turning the reverse-ID test from a model self-check into a reader-facing evaluation.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper proposes a multi-phase LLM workflow to characterize the investigative methods of seven fictional detectives: 15 LLMs independently generate short descriptions; GPT-4o extracts candidate traits and semantically groups them; traits are filtered by cross-model consensus at a 20% threshold; and the resulting profiles are evaluated by asking the same 15 LLMs to identify the detective from the trait list. The authors report 91.43% overall reverse-identification accuracy and a qualitative alignment of the trait profiles with literary criticism, interpreting these as evidence that the method reliably captures distinctive investigative approaches for reuse in narrative generation.

Significance. If the central claim held, the paper would supply a scalable, reusable protocol for character profiling in computational narratology, with a clearly documented prompt set, a concrete pipeline, and full trait lists in the appendix. The authors are transparent about their prompts, model versions, temperature settings, and the confusion matrix, which supports reproducibility. The main weakness is that the headline validation is circular: the same models that generate the descriptions are the ones asked to recognize them, and the literary comparison is post-hoc and qualitative. These issues are load-bearing for the claim that the profiles capture externally valid, distinctive investigative methods, so the manuscript needs additional validation before the effectiveness claim can be accepted.

major comments (3)
  1. [Section 3.6 and Section 4.2, Prompt 4 and Table 2] The reverse-identification test uses the same 15 LLMs that generated the descriptions in Section 3.2, with GPT-4o also performing trait extraction (Section 3.3) and semantic grouping (Section 3.4). High accuracy therefore largely measures self-consistency within the ensemble: the models are recognizing trait lists derived from their own outputs. The phrase in Section 4.2 that this 'demonstrat[es] the method's effectiveness' overstates what the test can establish. This is not an independent validation of distinctiveness or external accuracy. I recommend adding a held-out evaluator (e.g., human raters or a different set of models not involved in generation), a non-informative control condition, or an external ground-truth comparison to support the claim.
  2. [Section 3.5, consistency threshold] The 20% consistency threshold is introduced without justification and is quite permissive: a trait is retained if only 3 of the 15 models independently contributed it to the group. Because this threshold directly determines which traits enter the final profiles and thus affects the downstream reverse-identification results, the paper should provide a sensitivity analysis showing how the threshold affects both the synthesized profiles and the reported accuracy. Without such an analysis, the central quantitative result depends on an arbitrary free parameter.
  3. [Section 4.1, literary comparison] The claimed 'validation against existing literary analyses' is a qualitative, post-hoc alignment with selected sources rather than a systematic evaluation. For each detective, the paper finds references that match the extracted traits, including sources likely present in the LLMs' training data, so agreement with these references does not distinguish a genuine extraction from memorized common critical commentary. A pre-registered or independent-annotation protocol (e.g., human raters judging whether each trait is distinctive and accurate for the detective) would be needed to support the validation claim.
minor comments (4)
  1. [Abstract and Section 4.1] The abstract and Section 4.1 state that traits were 'validated against existing literary analyses,' but the paper presents no formal validation metric or protocol for that comparison; consider softening the wording to 'compared with' or 'consistent with selected literary analyses.'
  2. [Section 3.6] The phrase 'indirect evaluation of the clarity, uniqueness, and informativeness' attributes three properties to the profiles, but the reverse-identification task only directly tests whether the models can map the trait list to the correct name; uniqueness and informativeness are not separately measured.
  3. [Section 4.2] There is a typo in 'the 15 LLMs presented in Section3.6' — a missing space before the section number.
  4. [References] Reference [2] is incomplete: the author field appears as 'T. B. and' with no full name or title for the first author; this should be corrected.

Circularity Check

2 steps flagged · score 6.0 of 10

The 91.43% reverse-identification accuracy is a self-consistency result: the same 15 LLMs that produced the trait descriptions are the evaluators, so the central effectiveness claim is not externally validated.

  1. fitted input called prediction [Section 3.6 (Validation via Reverse Identification), Section 4.2 (Accuracy Analysis), Prompt 4]
    "The reverse identification task was performed using the same set of 15 LLMs employed in earlier phases (see Table 1). ... Model predictions were compared to the ground-truth detective associated with each trait list ... The results of this evaluation, presented in Table 2, demonstrate a high overall accuracy of 91.43%, with 96 out of 105 cases correctly identified. This high accuracy supports the effectiveness of our approach."

    The trait profiles are synthesized from descriptions generated by these same 15 LLMs in Phase 1 (Description Generation), then the 'validation' asks exactly those models to name the detective from the trait list, with the seven detective names supplied in Prompt 4. High agreement therefore mainly shows that the pipeline preserves the models' prior consensus about well-known detectives, not that the traits are externally accurate or distinctive. No held-out evaluator, non-informative control, or human/external ground truth is used, so the 91.43% figure is a self-consistency statistic presented as evidence of effectiveness.

  2. fitted input called prediction [Sections 3.3, 3.4 and 3.6]
    "For this phase, a single high-performing LLM (OpenAI's GPT-4o) was used to extract concise and non-redundant lists of traits from each description. ... The semantic grouping was performed using OpenAI's GPT-4o ... The reverse identification task was performed using the same set of 15 LLMs employed in earlier phases."

    GPT-4o is the sole model that converts the raw descriptions into trait lists and then groups those traits into the synthesized profiles; it is also one of the 15 models later asked to identify the detective from those profiles. This compounds the closed loop: the central processor of the trait data is also one of the evaluators, so part of the reported accuracy measures GPT-4o's agreement with its own extraction and grouping output rather than independent verification.

full rationale

The paper's central effectiveness claim rests on reverse identification: 'the identified traits were validated against existing literary analyses and further tested in a reverse identification phase, achieving an overall accuracy of 91.43%, demonstrating the method's effectiveness in capturing the distinctive investigative approaches of each detective.' That demonstration is self-referential by construction. The descriptions (Phase 1) come from the same 15 LLMs that later perform identification (Phase 3.6), and the trait extraction and semantic grouping (Phases 2 and 3) are performed by GPT-4o, which is itself one of those 15 evaluators. The test is a seven-way forced choice among extremely famous detectives whose methods are heavily represented in the models' training data, with the names provided in Prompt 4. High agreement therefore shows ensemble self-consistency, not externally validated accuracy. The comparison to literary studies in Section 4.1 is qualitative and post hoc, selecting references that broadly match the generated traits; it does not provide an independent quantitative gold standard. This is not a fully formal tautology, because the pipeline includes transformations and the accuracy is not mathematically forced, so a score of 6 (partial circularity) is appropriate rather than 8 or 10. The paper is otherwise self-contained in its methodology; the circularity is localized to the validation strategy for the main claim.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on assumptions that LLM consensus reflects true character traits, that GPT-4o's semantic grouping is lossless, and that the cited literary analyses are valid ground truth. No external corpus or human annotation is used to break the loop.

free parameters (2)
  • Consistency threshold = 20% (at least 3 of 15 LLMs)
    Chosen in Section 3.5 to filter traits; no principled justification or sensitivity analysis is given.
  • Maximum sentence limit in generation prompt = 5 sentences
    Structural constraint in Prompt 1 that shapes description length and content; hand-chosen.
assumptions (4)
  • domain assumption LLMs' parametric knowledge of the seven detectives is sufficiently accurate and representative of the characters' canonical investigative methods.
    The entire trait extraction relies on LLMs recalling accurate facts about these fictional characters from training data; no external corpus is used.
  • domain assumption Consensus across at least 3 LLMs indicates a true distinguishing trait rather than a common stereotype or training-data artifact.
    The 20% threshold in Section 3.5 operationalizes this assumption.
  • domain assumption GPT-4o's semantic grouping faithfully preserves the meaning of the extracted traits without introducing or losing information.
    Semantic grouping in Section 3.4 is a single-model transformation not separately validated.
  • domain assumption The cited literary analyses correctly describe the detectives' methods and can serve as ground truth.
    Section 4.1 compares identified traits to these analyses; the analyses themselves are taken as authoritative without independent confirmation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Characterizing the Investigative Methods of Fictional Detectives with Large Language Models." pith.science (2026). https://pith.science/paper/3QN3RMBK

@misc{pith2026250507601,
  author       = {Pith},
  title        = {Pith review of: Characterizing the Investigative Methods of Fictional Detectives with Large Language Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3QN3RMBK}},
  note         = {Machine review of arXiv:2505.07601}
}
read the original abstract

Detective fiction, a genre defined by its complex narrative structures and character-driven storytelling, presents unique challenges for computational narratology, a research field focused on integrating literary theory into automated narrative generation. While traditional literary studies have offered deep insights into the methods and archetypes of fictional detectives, these analyses often focus on a limited number of characters and lack the scalability needed for the extraction of unique traits that can be used to guide narrative generation methods. In this paper, we present an AI-driven approach for systematically characterizing the investigative methods of fictional detectives. Our multi-phase workflow explores the capabilities of 15 Large Language Models (LLMs) to extract, synthesize, and validate distinctive investigative traits of fictional detectives. This approach was tested on a diverse set of seven iconic detectives - Hercule Poirot, Sherlock Holmes, William Murdoch, Columbo, Father Brown, Miss Marple, and Auguste Dupin - capturing the distinctive investigative styles that define each character. The identified traits were validated against existing literary analyses and further tested in a reverse identification phase, achieving an overall accuracy of 91.43%, demonstrating the method's effectiveness in capturing the distinctive investigative approaches of each detective. This work contributes to the broader field of computational narratology by providing a scalable framework for character analysis, with potential applications in AI-driven interactive storytelling and automated narrative generation.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. CaughtCheating: Is Your MLLM a Good Cheating Detective? Exploring the Boundary of Visual Perception and Reasoning

    cs.CV 2025-06 conditional novelty 6.0 of 10

    A new 100-image benchmark shows that state-of-the-art multimodal models detect subtle, socially meaningful visual clues at near-chance levels and hallucinate accusations on innocent images.

Reference graph

Works this paper leans on

51 extracted references · 43 canonical work pages · cited by 1 Pith paper

  1. [1]

    Alabdulkarim, W

    A. Alabdulkarim, W. Li, L. J. Martin, and M. O. Riedl. Goal-directed story generation: Augmenting generative language models with reinforcement learning, 2021. URL https://arxiv.org/abs/2112.08593

    arXiv 2021

  2. [2]

    T. B. and. The rationalism of father brown. Perspectives on Political Science, 34(1):37–45, 2005. doi: 10.3200/ PPSC.34.1.37-45. 11 Characterizing the Investigative Methods of Fictional Detectives with Large Language Models

    work page 2005

  3. [3]

    T. J. Anderson, D. A. Schum, and W. Twining.Analysis of Evidence. Cambridge University Press, Cambridge, 2nd edition, 2005

    work page 2005

  4. [4]

    M. Bal. Narratology: Introduction to the Theory of Narrative. University of Toronto Press, 2007

    work page 2007

  5. [5]

    S. D. Barbosa, E. S. de Lima, A. L. Furtado, and B. Feijó. Generation and dramatization of detective stories. Journal on Interactive Systems, 5(2), Nov. 2014. doi: 10.5753/jis.2014.648

    work page doi:10.5753/jis.2014.648 2014

  6. [6]

    E. F. Bargainnier. The Gentle Art of Murder: The Detective Fiction of Agatha Christie. Bowling Green University Popular Press, 1980

    work page 1980

  7. [7]

    Berzsenyi

    C. Berzsenyi. Columbo: A Rhetoric of Inquiry with Resistant Responders. Intellect Ltd, 2021

    work page 2021

  8. [8]

    C. E. Berzsenyi. The columbo method - a legacy of antipotency and rhetorical inquiry to redirect resistance. Journal of Professional Communication, 6(2):7–36, 2021. doi: 10.15173/jpc.v6i2.4121

    work page doi:10.15173/jpc.v6i2.4121 2021

Show all 51 references
  1. [9]

    D. Carson. The abduction of sherlock holmes. International Journal of Police Science & Management, 11(2): 193–202, 2009. doi: 10.1350/ijps.2009.11.2.123

    2009 doi

  2. [10]

    Castricato, S

    L. Castricato, S. Frazier, J. Balloch, and M. Riedl. Tell me a story like i’m five: Story generation via question answering. In Proceedings of the 3rd Workshop on Narrative Understanding, 2021. URL https://par.nsf. gov/biblio/10249509

    2021

  3. [11]

    Cavazza and D

    M. Cavazza and D. Pizzi. Narratology for interactive storytelling: A critical introduction. In S. Göbel, R. Malkewitz, and I. Iurgel, editors, Technologies for Interactive Digital Storytelling and Entertainment, pages 72–83, Berlin, Heidelberg, 2006. Springer Berlin Heidelberg...

    2006 doi

  4. [12]

    E. S. de Lima, B. Feijó, and A. L. Furtado. Computational narrative blending based on planning. In J. Baal- srud Hauge, J. C. S. Cardoso, L. Roque, and P. A. Gonzalez-Calero, editors,Entertainment Computing – ICEC 2021, pages 89–303, Cham, 2021. Springer International Publishi...

    2021 doi

  5. [13]

    E. S. de Lima, B. Feijó, and A. L. Furtado. Procedural generation of branching quests for games. Entertainment Computing, 43:100491, 2022. ISSN 1875–9521. doi: 10.1016/j.entcom.2022.100491

    2022

  6. [14]

    E. S. de Lima, M. A. Casanova, B. Feijó, and A. L. Furtado. Semiotic structuring in movie narrative generation. In P. Ciancarini, A. Di Iorio, H. Hlavacs, and F. Poggi, editors,Entertainment Computing – ICEC 2023, pages 161–175, Singapore, 2023. Springer Nature Singapore. doi:...

    2023 doi

  7. [15]

    E. S. de Lima, B. Feijó, and A. L. Furtado. Managing the plot structure of character-based interactive narratives in games. Entertainment Computing, 47:100590, 2023. ISSN 1875–9521. doi: 10.1016/j.entcom.2023.100590

    2023

  8. [16]

    E. S. de Lima, M. A. Casanova, and A. L. Furtado. Imagining from images with an AI storytelling tool, 2024. URL https://arxiv.org/abs/2408.11517

    2024 arXiv

  9. [17]

    E. S. de Lima, B. Feijó, M. A. Casanova, and A. L. Furtado. Chatgeppetto - an AI-powered storyteller. In Proceedings of the 22nd Brazilian Symposium on Games and Digital Entertainment, pages 28–37. ACM, 2024. doi: 10.1145/3631085.3631302

    2024

  10. [18]

    E. S. de Lima, M. M. E. Neggers, M. A. Casanova, B. Feijó, and A. L. Furtado. A pattern-oriented AI-powered approach to story composition. In P. Figueroa, A. Di Iorio, D. Guzman del Rio, E. W. Gonzalez Clua, and L. Cuevas Rodriguez, editors, Entertainment Computing – ICEC 2024...

    2024 doi

  11. [19]

    E. S. de Lima, M. M. Neggers, M. A. Casanova, and A. L. Furtado. From images to stories: Exploring player- driven narratives in games. In A. Marto, R. Prada, P. Gouveia, R. Contreras-Espinosa, A. Gonçalves, E. Abrantes, and R. Ribeiro, editors, Videogame Sciences and Arts, pag...

    2025 doi

  12. [20]

    E. S. de Lima, M. M. Neggers, B. Feijó, M. A. Casanova, and A. L. Furtado. An AI-powered approach to the semiotic reconstruction of narratives. Entertainment Computing, 52:100810, 2025. doi: 10.1016/j.entcom.2024. 100810

    2025 doi

  13. [21]

    E. S. de Lima, M. M. E. Neggers, and A. L. Furtado. Multigenre AI-powered story composition, 2025. URL https://arxiv.org/abs/2405.06685

    2025 arXiv

  14. [22]

    Domjanovi´c

    V . Domjanovi´c. The comparative analysis of murder and crime-solving strategies in agatha christie’s works. Undergraduate thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Humanities and Social Sciences,

  15. [23]

    Domjanovi´c and B

    V . Domjanovi´c and B. Oklop ˇci´c. Murder and crime-solving strategies in agatha christie’s works. Journal of Language and Literary Studies, XIII(39):9–29, February 2022. doi: 10.31902/fll.39.2022.1. 12 Characterizing the Investigative Methods of Fictional Detectives with Lar...

    2022 doi

  16. [24]

    Eco and T

    U. Eco and T. A. Sebeok, editors. The Sign of the Three: Dupin, Holmes, Peirce . Indiana University Press, Bloomington, 1988

    1988

  17. [25]

    A. Fan, M. Lewis, and Y . Dauphin. Hierarchical neural story generation. In I. Gurevych and Y . Miyao, editors, Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 889–898, Melbourne, Australia, 2018. Associati...

    2018 doi

  18. [26]

    Feijó, E

    B. Feijó, E. S. de Lima, and A. L. Furtado. A transdisciplinary approach to computational narratology. In Proceedings of the XX Brazilian Symposium on Computer Games and Digital Entertainment (SBGames 2021), Culture Track, Gramado, Brazil, October 2021

    2021

  19. [27]

    R. Friedel. Serendipity is no accident. The Kenyon Review, 23(2):36–47, 2001. URL http://www.jstor.org/ stable/4338198

    2001

  20. [28]

    N. Frye. Anatomy of Criticism: Four Essays . Princeton Classics. Princeton University Press, Princeton, 2nd edition, 2020

    2020

  21. [29]

    G. Grella. Murder and manners: The formal detective novel. NOVEL: A Forum on Fiction, 4(1):30–48, 1970. doi: 10.2307/1345250

    1970 doi

  22. [30]

    Havlí ˇcková

    A. Havlí ˇcková. Agatha christie and her great detective (based on poirot investigates and hercule poirot’s christmas, reflecting 1920s and 1930s). Bachelor’s thesis, Masaryk University, Faculty of Arts, 2005. URL https://is.muni.cz/th/cwjlr/

    2005

  23. [31]

    M. Jenner. ’Follow the Evidence’? Methods of Detection in American TV Detective Drama. Phd thesis, Department of Theatre, Film and Television Studies, Aberystwyth University, October 2013

    2013

  24. [32]

    Ketovi´c

    M. Ketovi´c. Detective archetypes in anglophone fiction. Master’s thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Humanities and Social Sciences, 2019. URL https://urn.nsk.hr/urn:nbn:hr:142: 968848

    2019

  25. [33]

    E. Laven. Detection, desire and contamination: The strange case of sherlock holmes. Master thesis, University of Waikato, Hamilton, New Zealand, 2013. URL https://hdl.handle.net/10289/8466

    2013

  26. [34]

    S. J. Link. Defining Detective Fiction, pages 17–38. Springer Nature Switzerland, Cham, 2023. doi: 10.1007/ 978-3-031-33227-2_2

    2023

  27. [35]

    J. Lu, Z. Pang, M. Xiao, Y . Zhu, R. Xia, and J. Zhang. Merge, ensemble, and cooperate! a survey on collaborative strategies in the era of large language models, 2024. URL https://arxiv.org/abs/2407.06089

    2024 arXiv

  28. [36]

    Michalowicz

    N. Michalowicz. ‘magnificent intellect’: Character, intelligence, and genius in sherlock holmes. Victoriographies, 13(2):210–228, 2023. doi: 10.3366/vic.2023.0493

    2023

  29. [37]

    C. Monico. Agatha christie: A look into criminal procedure and gender. Thesis/capstone/creative project, Uni- versity of Nebraska at Omaha, 2021. URL https://digitalcommons.unomaha.edu/university_honors_ program/157

    2021

  30. [38]

    P. Murray. Classical Literary Criticism. Oxford University Press, Oxford, 2000

    2000

  31. [39]

    Oklopcic

    B. Oklopcic. The image of central european immigrant in popular fiction and its adaptations: A case study of the detective murdoch/murdoch mysteries series. European Journal of American Studies, 18(4), 2023. doi: 10.4000/ejas.20949

    2023 doi

  32. [40]

    Petrescu

    Z.-M. Petrescu. Stylistic use of the superiority theory in murdoch mysteries television series. In The 7th International Language, Culture and Literature Symposium Book of Proceedings, pages 79–87, December 2022. ISBN 978-605-73128-1-5

    2022

  33. [41]

    Rapezzi, R

    C. Rapezzi, R. Ferrari, and A. Branzi. White coats and fingerprints: Diagnostic reasoning in medicine and investigative methods of fictional detectives. BMJ (Clinical Research Ed.), 331(7531):1491–1494, 2005. doi: 10.1136/bmj.331.7531.1491

    2005

  34. [42]

    M.-L. Ryan. Interactive narrative, plot types, and interpersonal relations. In U. Spierling and N. Szilas, editors, Interactive Storytelling, pages 6–13, Berlin, Heidelberg, 2008. Springer Berlin Heidelberg

    2008

  35. [43]

    E. Sheley. Criminal investigation and canadian national identity in murdoch mysteries. Entertainment and Sports Law Journal, 18:1–6, 2020. doi: 10.16997/eslj.253

    2020 doi

  36. [44]

    T. Todorov. The Poetics of Prose. Cornell University Press, 1977

    1977

  37. [45]

    van Laethem

    E. van Laethem. Connecting detectives: Investigation of the extent to which arthur conan doyle was influenced by edgar allan poe’s detective auguste dupin when creating sherlock holmes. Master’s thesis, Ghent University, Faculty of Arts and Philosophy, 2017. 13 Characterizing ...

    2017

  38. [46]

    Värtinen, P

    S. Värtinen, P. Hämäläinen, and C. Guckelsberger. Generating role-playing game quests with gpt language models. IEEE Transactions on Games, 16(1):127–139, 2024. doi: 10.1109/TG.2022.3228480

    2024

  39. [47]

    H. Walpole. Letter to horace mann, january 28, 1754. Letter, 1754. URL https://archive.org/details/ lettersofhoracew01walp/page/408

  40. [48]

    Xie and M

    K. Xie and M. Riedl. Creating suspenseful stories: Iterative planning with large language models. In Y . Graham and M. Purver, editors, Proceedings of the 18th Conference of the European Chapter of the Association for Com- putational Linguistics (Volume 1: Long Papers), pages ...

    2024

  41. [49]

    P. Xu, M. Patwary, M. Shoeybi, R. Puri, P. Fung, A. Anandkumar, and B. Catanzaro. MEGATRON-CNTRL: Controllable story generation with external knowledge using large-scale language models. In B. Webber, T. Cohn, Y . He, and Y . Liu, editors,Proceedings of the 2020 Conference on ...

    2020

  42. [50]

    K. Yang, Y . Tian, N. Peng, and D. Klein. Re3: Generating longer stories with recursive reprompting and revision. In Y . Goldberg, Z. Kozareva, and Y . Zhang, editors,Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 4393–4479, Abu D...

    2022 doi

  43. [2019]

    URL https://urn.nsk.hr/urn:nbn:hr:142:488458

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.