Pith. sign in

REVIEW 3 major objections 4 minor 27 references

Sanremo Festival lyrics have grown steadily more semantically uniform over 75 years.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-11 12:48 UTC pith:WF3Q4K22

load-bearing objection Clean multi-level pipeline on a complete Sanremo corpus shows rising lyrical uniformity that matches English trends; the result is real but rests on unvalidated embedding proxies. the 3 major comments →

arxiv 2607.04832 v1 pith:WF3Q4K22 submitted 2026-07-06 cs.CL

Semantic Homogenization in Italian Popular Music: A Diachronic Analysis

Lorenzo Canale , Stefano Scotta , Alberto Messina This is my paper
classification cs.CL
keywords semantic similarityembedding modelssong lyrics analysislarge language modelsSanremo Festivaldiachronic analysislyrical homogenization
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper asks whether the well-documented decline in lyrical variety in English popular music also appears in a different language and cultural setting: the complete set of finalist songs from every Sanremo Music Festival since 1951. The authors build a multi-view pipeline that measures similarity four ways—full lyrics, key segments, Gemini-extracted topics, and significant-word overlap—using three modern embedding models plus large language models for segmentation and topic extraction. When the four views are averaged, the resulting year-by-year matrix shows a clear, gradual rise in semantic uniformity both within each festival edition and across editions, with the strongest homogenization in recent decades. The result matters because Sanremo is Italy’s central annual cultural stage; a measurable narrowing of its lyrical language is therefore a concrete signal of longer-term shifts in national musical expression.

Core claim

Across full-text, portion-based, topic-based and word-based similarity matrices (averaged over three embedding models), the Sanremo corpus exhibits a gradual increase in semantic uniformity both within and across years, most pronounced in recent decades.

What carries the argument

The composite year-by-year matrix Z, formed by summing and normalising the four independent similarity views (full-text, top-k portions, topics, word intersections) so that each pair of festival years receives a single scalar that aggregates every measurement.

Load-bearing premise

That cosine similarity of the chosen multilingual embeddings and of Gemini-extracted topics reliably tracks what Italian listeners would judge as semantic similarity of song lyrics, without any direct human validation on the Sanremo material itself.

What would settle it

A controlled human rating study in which Italian listeners score pairwise lyrical similarity for a stratified sample of Sanremo song pairs; if the correlation between those scores and the paper’s matrix Z entries is near zero, the central claim collapses.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper develops a multi-route methodology (full-text cosine/MaxSim embeddings, top-k portion similarity, LLM-extracted topic embeddings, and word-overlap after PoS/normalization) using three embedding models plus Gemini-1.5-flash, then applies it to the complete set of Sanremo finalist lyrics (1951–2025). Averaging the resulting year-pair matrices into a composite Z (Section 7) yields a visual pattern of rising within-year and cross-year semantic similarity, most pronounced in recent decades, which the authors interpret as progressive lyrical homogenization consistent with English-language findings.

Significance. If the observed trend is genuine, the work supplies a culturally specific Italian counterpart to recent English-lyric simplification studies and offers a reusable, multi-granularity pipeline that other researchers can apply to any time-stamped lyric corpus. The high Pearson correlations among four independent analysis routes (Figure 7) and the public release of song metadata constitute concrete strengths; the methodology itself is a useful contribution even if the Sanremo-specific claim requires further validation.

major comments (3)
  1. [Section 7, Figures 8–9] Section 7 and Figures 8–9 present the central claim of increasing semantic uniformity solely via visual inspection of the composite matrix Z and its diagonal/row-mean time series. No regression, Mann–Kendall test, or other formal assessment of the temporal trend is reported, nor are confidence intervals or bootstrap uncertainty estimates supplied for the matrix entries. Without such quantification the claimed “gradual move toward increasing semantic uniformity” remains an informal observation rather than a statistically supported result.
  2. [Section 3, Section 5, Section 6] The entire pipeline rests on the untested assumption that cosine (or symmetrized MaxSim) similarities of multilingual-e5-large, text-embedding-3-large and ColBERT-XM vectors, together with Gemini-extracted topics and portions, faithfully track human-perceived semantic similarity of Italian song lyrics. Section 3 cites only a single external correlation study that used a different English model (all-MiniLM-L6-v2). No human similarity ratings, inter-annotator agreement, or even a small Italian pilot on the Sanremo material itself are provided. This missing external anchor is load-bearing for the interpretation of warmer colors in recent decades of Z.
  3. [Section 6.2 Remark 1, Section 7] Free parameters (k=3 for top-k portions, q=0.75 quantile threshold, equal-weight averaging across embeddings and analysis types) are fixed without sensitivity analysis or justification beyond “arbitrary small” (Remark 1). Because the composite Z is formed by summing these choices, it is unclear how robust the reported homogenization pattern is to alternative parameter settings.
minor comments (4)
  1. [Declarations / Data availability] Data-availability statement notes that only metadata (year, title, author, interpreter) are released because of copyright. While understandable, this prevents direct reproduction of the embedding matrices; the authors should at least deposit the year-pair similarity matrices themselves.
  2. [Section 6.4, Section 7] Notation for the four word-based matrices W^ι is introduced in Table 1 but the subsequent averaging step that produces W_mean is described only in prose; an explicit equation would improve clarity.
  3. [Figure 1] Figure 1 caption is long and dense; splitting the methodological overview from the concrete 1956–2004 example would aid readability.
  4. [References] Several references appear as arXiv preprints without final venue information; updating them where possible would strengthen the bibliography.

Circularity Check

0 steps flagged

No circularity: purely empirical multi-method similarity analysis on an external corpus with no fitted parameters, self-definitional quantities, or load-bearing self-citations.

full rationale

The paper's central claim (gradual rise in semantic uniformity of Sanremo lyrics, visible in composite matrix Z and its diagonal) is an observational result obtained by applying four independent similarity pipelines (full-text cosine/MaxSim of three embedding models, top-k portion similarities, Gemini-extracted topic embeddings, and statistical word-overlap after LLM normalization) to a fixed external corpus of festival finalists, then averaging the resulting year-pair matrices. No quantity is defined in terms of another that is later 'predicted'; no parameters are fitted to a subset and then used to forecast a related quantity; no uniqueness theorem or ansatz is imported via self-citation; and the high Pearson correlations among the four pipelines (Fig. 7) are post-hoc consistency checks rather than circular constructions. The methodology is self-contained against the Sanremo data and does not reduce by construction to its inputs. Minor external citations (e.g., to an English MiniLM human-correlation study) are not load-bearing for the Italian result. Score 0 is therefore the correct, non-manufactured finding.

Axiom & Free-Parameter Ledger

3 free parameters · 3 axioms · 0 invented entities

The central claim rests on the assumption that off-the-shelf multilingual embeddings plus Gemini topic extraction constitute a valid semantic similarity metric for Italian lyrics, plus a handful of free analysis choices (k, quantile, averaging). No new physical or mathematical entities are postulated; the free parameters are purely methodological knobs.

free parameters (3)
  • k (top-k portions) = 3
    Fixed at 3 without sensitivity analysis; controls how many highest-similarity segments contribute to portion-based matrices (Section 6.2).
  • quantile threshold q = 0.75
    Set to 0.75 for the proportion-of-high-similarity matrices; arbitrary high-quantile choice (Section 6.1).
  • equal-weight average across embeddings and analysis types = uniform sum
    F_mean, P_mean, T_mean, W_mean are simply summed to form Z; no learned or justified weights (Section 7).
axioms (3)
  • domain assumption Cosine (or symmetrized MaxSim) similarity of the chosen embedding models is a faithful proxy for human semantic similarity of song lyrics.
    Invoked throughout Section 6; supported only by a citation to a different English model (Kim & Akama 2024).
  • domain assumption Gemini-1.5-flash produces reliable topic lists and lyric segmentations for Italian songs.
    Used without validation or inter-annotator agreement in Sections 6.2–6.3.
  • domain assumption Finalist songs of Sanremo constitute a representative sample of Italian popular-music lyrical evolution.
    Implicit in the claim that the observed trend generalizes to ‘Italian popular music’ (Abstract, Introduction).

pith-pipeline@v1.1.0-grok45 · 15409 in / 2488 out tokens · 23209 ms · 2026-07-11T12:48:25.738590+00:00 · methodology

0 comments
read the original abstract

In recent years, studies have revealed a decline in semantic variety across popular music lyrics, particularly in English-language songs on streaming platforms like Spotify. This research examines whether a similar trend can be observed in a different linguistic and cultural context: the lyrics of all finalist songs from the 75 editions of the Sanremo Music Festival, Italy's most renowned music competition. What sets this work apart is the development of a flexible and efficient methodology for tracking changes in semantic similarity over time, which can be applied to different datasets to study similar phenomena. Drawing on a combination of full-text, segment-based, topic-based, and word-level analyses, the approach leverages both embedding techniques and large language models. When applied to the Sanremo corpus, this framework reveals a gradual move toward increasing semantic uniformity, echoing the global patterns identified in previous studies. These findings underscore the value of natural language processing tools in uncovering long-term shifts in musical language and cultural expression.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

27 extracted references · 9 canonical work pages · 1 internal anchor

  1. [1]

    Elsevier, Netherlands (2009)

    Allan, K.: Concise Encyclopedia of Semantics. Elsevier, Netherlands (2009)

    2009

  2. [2]

    Journal of Literary Semantics41(1), 47–65 (2012) https://doi.org/10.1515/jls-2012-0003

    Panagiotidou, M.-E.: An introduction to the semantics of intertextuality. Journal of Literary Semantics41(1), 47–65 (2012) https://doi.org/10.1515/jls-2012-0003

    work page doi:10.1515/jls-2012-0003 2012

  3. [3]

    Springer, Cham (2022)

    Harispe, S., Ranwez, S., Janaqi, S., Montmain, J.: Semantic Similarity from Nat- ural Language and Ontology Analysis. Springer, Cham (2022). https://doi.org/ 10.1007/978-3-031-02156-5

    doi:10.1007/978-3-031-02156-5 2022

  4. [4]

    In: Dietrich, E

    Rapaport, W.J.: Syntactic semantics: Foundations of computational natural- language understanding. In: Dietrich, E. (ed.) Thinking Computers and Virtual Persons, pp. 225–273. Academic Press, University of Michigan (1994). https: //doi.org/10.1016/B978-0-12-215495-9.50015-6

    work page doi:10.1016/b978-0-12-215495-9.50015-6 1994

  5. [5]

    https://doi.org/10.18653/v1/2020

    Li, B., Zhou, H., He, J., Wang, M., Yang, Y., Li, L.: On the Sentence Embed- dings from Pre-trained Language Models. https://doi.org/10.18653/v1/2020. emnlp-main.733

    doi:10.18653/v1/2020 2020

  6. [6]

    Proceedings of Workshop at ICLR2013(2013)

    Mikolov, T., Chen, K., Corrado, G., Dean, J.: Efficient estimation of word representations in vector space. Proceedings of Workshop at ICLR2013(2013)

    2013

  7. [7]

    EMNLP14, 1532–1543 (2014) https://doi.org/10.3115/v1/D14-1162

    Pennington, J., Socher, R., Manning, C.: Glove: Global vectors for word repre- sentation. EMNLP14, 1532–1543 (2014) https://doi.org/10.3115/v1/D14-1162

    doi:10.3115/v1/d14-1162 2014

  8. [8]

    https://doi

    Kalyan, K.S., Rajasekharan, A., Sangeetha, S.: Ammus: A Survey of Transformer- Based Pretrained Models in Natural Language Processing (2021). https://doi. org/10.48550/arXiv.2108.05542

    Pith/arXiv arXiv doi:10.48550/arxiv.2108.05542 2021

  9. [9]

    Parole e Suoni Raccontano la Nazione

    Facci, S.: Il Festival di Sanremo. Parole e Suoni Raccontano la Nazione. Carocci Editore, Rome, Italy (2011) 16

    2011

  10. [10]

    Popular Music26(3), 389–408 (2007)

    Agostini, R.: The italian “canzone” and the sanremo festival: Change and con- tinuity in italian mainstream pop of the 1960s. Popular Music26(3), 389–408 (2007). Accessed 2025-05-06

    2007

  11. [11]

    https://scholarwolf.unr.edu/bitstreams/ db60b3d3-538b-485d-b877-f3e50be924b7/download

    Pierce, C.A.: Shifting Boundaries: Exploring Queer and National Iden- tity in the Sanremo Festival (2023). https://scholarwolf.unr.edu/bitstreams/ db60b3d3-538b-485d-b877-f3e50be924b7/download

    2023

  12. [12]

    Communication, Culture and Critique13(1), 1–16 (2020) https://doi.org/10

    Ardizzoni, M.: On rhythms and rhymes: Poetics of identity in postcolonial italy. Communication, Culture and Critique13(1), 1–16 (2020) https://doi.org/10. 1093/ccc/tcz049

    2020

  13. [13]

    Italianness

    Caiani, M., Padoan, E.: Playing “Italianness” in Popular Music. In: Popular Music and the Rise of Populism In Europe, 1st edn., pp. 58–83. Routledge, London (2024). https://doi.org/10.4324/9781003293156-3

    work page doi:10.4324/9781003293156-3 2024

  14. [14]

    The Sanremo Festival and the Cir- culation of Populist Discourses

    Magaudda, P.: Populism, Music and the Media. The Sanremo Festival and the Cir- culation of Populist Discourses. Partecipazione e conflitto13(1), 132–153 (2020) https://doi.org/10.1285/I20356609V13I1P132

    work page doi:10.1285/i20356609v13i1p132 2020

  15. [15]

    https://evidencenetwork.ca/ sanremo-seen-from-the-bed-psychology-among-the-scores/

    Muller, D.K.: Sanremo seen from the bed: psychol- ogy among the scores (2025). https://evidencenetwork.ca/ sanremo-seen-from-the-bed-psychology-among-the-scores/

    2025

  16. [16]

    https://doi.org/10.48550/ arXiv.2501.07557

    Marco, N.D., Loru, E., Galeazzi, A., Cinelli, M., Quattrociocchi, W.: Decoding Musical Evolution Through Network Science (2025). https://doi.org/10.48550/ arXiv.2501.07557

    Pith/arXiv arXiv 2025

  17. [17]

    charts from 1950 to 2023

    Hamilton, M., Pearce, M.: Trajectories and revolutions in popular melody based on u.s. charts from 1950 to 2023. Scientific Reports14(2024) https://doi.org/ 10.1038/s41598-024-64571-x

    work page doi:10.1038/s41598-024-64571-x 1950

  18. [18]

    In: 2020 International Conference on Emerging Trends in Information Technology and Engineering (ic-ETITE), pp

    Chandra, J., Santhanam, A., Joseph, A.: Artificial intelligence based semantic text similarity for rap lyrics. In: 2020 International Conference on Emerging Trends in Information Technology and Engineering (ic-ETITE), pp. 1–5 (2020). https: //doi.org/10.1109/ic-ETITE47903.2020.211

    work page doi:10.1109/ic-etite47903.2020.211 2020

  19. [19]

    In: 2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA), pp

    Ogihara, M., Galarraga, D., Ren, G., Tavares, T.: The semantic shapes of popular music lyrics: Graph-based representation, analysis, and interpretation of popular music lyrics in semantic natural language embedding space. In: 2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA), pp. 1249–1254 (2018). https://doi.org/10.1109...

    work page doi:10.1109/icmla.2018.00203 2018

  20. [20]

    A Computational Analysis of Lyric Similarity Perception

    Kim, H., Akama, T.: A Computational Analysis of Lyric Similarity Perception (2024). https://doi.org/10.48550/arXiv.2404.02342

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2404.02342 2024

  21. [21]

    https://doi.org/10.48550/arXiv.2502.13595

    Enevoldsen, K., Chung, I., Kerboua, I., Kardos, M., Mathur, A., Stap, D., Gala, 17 J., Siblini, W., Krzemi´ nski, D., Winata, G.I., Sturua, S., Utpala, S., Ciancone, M., Schaeffer, M., Sequeira, G., Misra, D., Dhakal, S., Rystrøm, J., Solomatin, R., ..., Muennighoff, N.: MMTEB: Massive Multilingual Text Embedding Benchmark (2025). https://doi.org/10.48550...

    doi:10.48550/arxiv.2502.13595 2025

  22. [22]

    on the multi-dimensional differences between german poetry of realism and modernism

    Ehrmanntraut, A., Hagen, T., Jannidis, F., Konle, L., Kr¨ oncke, M., Winko, S.: Modeling and measuring short text similarities. on the multi-dimensional differences between german poetry of realism and modernism. Journal of Com- putational Literary Studies1(1), 1–30 (2022) https://doi.org/10.48694/jcls.116

    work page doi:10.48694/jcls.116 2022

  23. [23]

    Scientific Reports14(1), 5531 (2024) https://doi.org/10.1038/ s41598-024-55742-x

    Parada-Cabaleiro, E., Mayerl, M., Brandl, S., Skowron, M., Schedl, M., Lex, E., Zangerle, E.: Song lyrics have become simpler and more repetitive over the last five decades. Scientific Reports14(1), 5531 (2024) https://doi.org/10.1038/ s41598-024-55742-x

    2024

  24. [24]

    https://doi.org/10.48550/arXiv

    Wang, L., Yang, N., Huang, X., Yang, L., Majumder, R., Wei, F.: Multilingual E5 Text Embeddings: A Technical Report (2024). https://doi.org/10.48550/arXiv. 2402.05672

    Pith/arXiv arXiv doi:10.48550/arxiv 2024

  25. [25]

    https://doi.org/10.48550/ arXiv.2004.12832

    Khattab, O., Zaharia, M.: ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT (2020). https://doi.org/10.48550/ arXiv.2004.12832

    Pith/arXiv arXiv 2020

  26. [26]

    https://doi.org/10.48550/arXiv.2112.01488

    Santhanam, K., Khattab, O., Saad-Falcon, J., Potts, C., Zaharia, M.: Col- BERTv2: Effective and Efficient Retrieval via Lightweight Late Interaction (2021). https://doi.org/10.48550/arXiv.2112.01488

    Pith/arXiv arXiv doi:10.48550/arxiv.2112.01488 2021

  27. [27]

    https://doi.org/10.5281/zenodo

    Honnibal, M., Ines, I.M., Sofie, S.V.L., Adriane, A.B.: spaCy: Industrial-strength Natural Language Processing in Python (2020). https://doi.org/10.5281/zenodo. 1212303 18

    doi:10.5281/zenodo 2020