Birds / Behavioral experiment

Japanese-tit call order: evidence and interpretations

Source year 2016 · Note reviewed 2026-10-14

Japanese great tits showed different receiver behaviors under component calls, natural versus reversed order, and single- versus two-source playback; a bounded novel cross-species sequence also elicited stronger responses than its reversal, while a matched ABC–zi control was weak. In the 2022 deposited binary table, the natural single-source 1A-R condition had an approach event in 13/16 flock trials and wing-flicking in 15/16; the other three treatments had approach events in 2–3/16 and wing-flicking in 1–4/16. These aggregate counts reproduce the descriptive pattern only. They do not establish a unique mechanism: one published critique agrees that receiver behavior is sensitive to order and source but disputes that this demonstrates Core-Merge or hierarchical syntax.

What we reviewed · Primary evidence and alternative theory reviewed

Direct publisher full text reviewed for the 2016 primary paper (Methods, Experiments 1–2, Results, figures and ethics) and 2022 follow-up (selected Methods, Results, ethics, data/code statements and rights). The linked 2022 Figshare v1 tables were checked against the current official API sizes and MD5 values, paired on trial metadata, and summarized with the project standard-library script; the deposited R scripts were read as text, not executed. For the 2017 paper, publisher-indexed primary text exposed the abstract, selected Results, STAR Methods, statistics and data statement; its direct ScienceDirect page returned HTTP 403. The 2017 Mendeley Data v1 landing metadata and record-level license were reviewed, but its files were not. For Schlenker et al. (2024), the main text (Sections I–VII) of the official HAL v1 author-accepted manuscript was reviewed; the formal appendices were inspected selectively, not recalculated. The separate Wiley supporting-information PDF returned HTTP 403 and was not reviewed. Beckers et al. (2024) Sections 1–4 were read in the full text available through PubMed Central; its conceptual critique is not a new animal experiment. For the adjacent Coye & Dutour (2025) study (source 268), publisher-indexed Methods, Results and Discussion passages and the University of Western Australia repository abstract/access record were reviewed; direct ScienceDirect retrieval returned HTTP 403, so the full article file, figures, supplement and complete statistical details were not reviewed. No inferential model was reproduced and no independent Talk2Nature specialist review has occurred.

Not yet independently reviewed by a scientific specialist.

Study at a glance

Study unit: 2016: 21 adult birds in the component experiment; 34 different birds in the order experiment (17 per treatment).
Design: 2016 component experiment: ABC, D, ABC–D and matched background control; call treatments used matched exemplars and randomized order.
Ordering test: 2016 order test: natural ABC–D versus reversed D–ABC, with call exemplars and interval features balanced; one treatment per bird.
Novel sequence: 2017: 56 birds, 14 per condition; ABC–tää responses were stronger than tää–ABC, while ABC–zi and zi–ABC responses were both weak and similar.
Call controls: 2017 follow-up: 42 flocks, 14 per call type; D, tää and shortened-tää playbacks tested call category and acoustic-similarity alternatives.
Source-linkage test: 2022 follow-up: 64 flocks, 16 per condition; one/two speakers crossed with natural/reversed order, with a shrike model present.
2022 deposit inspection: The Figshare v1 deposit has two paired 64-row behavioral outcome tables and two R scripts. A standard-library audit matched the table trial metadata and binary/count outcomes; the single-source natural 1A-R group had approach in 13/16 flocks and wing-flicking in 15/16. This is descriptive only; no mixed model or deposited code was run. No audio is present.
Adjacent cross-family comparator (2025): Coye & Dutour: 96 great tits, six treatments, 16 per treatment; approach did not differ among call treatments, while wing-flicking was more frequent after complete Carolina-wren or great-tit FME–D sequences than after D calls or controls. No reversed-order condition was tested.
2024 critique: Beckers et al. agree the experiments show order sensitivity and different responses to one-source versus two-source AR, but argue this does not establish Core-Merge or hierarchy; finite-list and finite-state/concatenation alternatives remain.
Outcomes: Horizontal scanning, approach and wing-flicking displays, not an AI-generated verbal interpretation. The 2024 theory papers add alternative explanations but no new playback result.

Reading location: Suzuki et al. 2016: Methods, Experiments 1–2; Results; Figures 3–4; ethics. Suzuki et al. 2017: publisher-indexed Results, STAR Methods, statistics and data availability; direct page inaccessible. Suzuki & Matsumoto 2022: selected Methods/Results; Figures 3–4; data/code availability; rights. Figshare item 18007046 v1: two tables paired and descriptively aggregated after current API size/MD5 verification; two R scripts read as text only. Beckers et al. 2024: Sections 1–4 of full text through PMC. Phillips & Wilson 2016: discussion of systematicity. Schlenker et al. 2024: main text Sections I–VII in the HAL v1 author-accepted manuscript; appendices inspected selectively, formal derivations not independently audited; publisher supplement not retrieved. Coye & Dutour (2025): publisher-indexed Methods, Results and Discussion passages; direct ScienceDirect retrieval returned HTTP 403. University of Western Australia repository abstract, publication and article-access/license metadata. No article file or supplementary material inspected.

Two explanations, one unresolved test

Schlenker et al. (2024) compare Minimal Compositionality and Bird Implicatures for ABC–D. Beckers et al. (2024) challenge the Core-Merge interpretation and discuss finite-list and concatenation-based finite-state accounts. These are published interpretations of prior receiver behavior, not competing trained models or new experiments.

Minimal Compositionality

Account: ABC–D combines the call meanings under one existential claim: the same noteworthy event licenses both alert and recruitment. Separate ABC and D utterances make two claims, which need not refer to the same event.

Distinctive prediction: Separate calls can still support mobbing if receivers can link both to the same threat, but the link should be harder than for a single ABC–D sequence.

Key assumption: D alone must carry an event-like, existential meaning. The authors note that this is not established by the cited approach response.

Bird Implicatures

Account: ABC–D is treated as a more informative alternative to ABC and/or D. A standalone call can imply that the fuller ABC–D combination is false, which can block a mobbing response even if the calls are understood as separate utterances.

Distinctive prediction: Source-separated ABC and D should tend toward a non-mobbing inference; this account does not require D alone to refer to a noteworthy event.

Key assumption: The analysis needs ABC–D to be evoked as an alternative. The authors favor a receiver-centered version to handle the hybrid ABC–D* result.

Finite-list and finite-state/concatenation alternatives (Beckers et al., 2024)

Account: A finite-list explanation treats A, R and the natural AR call as distinct complex vocalization types and can fit the 2022 treatment pattern without inferring a productive combinatorial operation. To accommodate order sensitivity across the 2016, 2017 and 2022 studies, the authors suggest a finite-state account based on ordered call bigrams.

Distinctive prediction: The cited results can show learned sensitivity to call identity, linear order and source while leaving hierarchical structure and open-ended productivity unestablished.

Key assumption: The critique is a conceptual analysis, not a new experiment or quantitative reanalysis. Its finite-list and finite-state explanations are alternatives to the interpretation, not directly tested models.

The 2022 result supports a bounded receiver-response difference between one-source natural AR and the other tested conditions. It does not uniquely establish Core-Merge, recursion or general language syntax. Schlenker et al. offer two formal accounts; Beckers et al. accept order sensitivity but argue finite-list or concatenation-based accounts can explain the evidence. The 2017 cross-species result remains a point any unified account must address.

What this does — and doesn’t — establish

These are species-, population- and task-bounded receiver-response results, not general animal-language or translation evidence. The 2017 novelty test combined a familiar heterospecific call type with Japanese-tit calls; it does not establish open-ended productivity, and its comparison with ABC–D used earlier data from another year. The 2022 study extends the same research line in the same species, not an independent replication. Its two Figshare v1 outcome tables contain no audio and no location fields; dates and caller_ID fields remain in ignored local files. The current official Figshare API reports MD5 values and the two received data tables match them; this confirms byte agreement with the current provider manifest, not scientific validity or a data-to-figure reproduction. The committed summary contains aggregates only. Its script does not run the deposited R code or reproduce mixed-effects models and adjusted contrasts. Schlenker et al. (2024) offer two formal accounts that fit the source-separation result under different assumptions. Beckers et al. (2024) agree the evidence shows sensitivity to call order and distinguishable responses to 1AR versus 2AR, but argue that the experiment does not establish Core-Merge or hierarchical syntax; a finite-list account fits the 2022 result, and a finite-state/concatenation account is proposed for the broader order pattern. This is a published interpretive disagreement, not a new empirical result or consensus. The 2025 Carolina-wren/great-tit study is adjacent cross-family receiver-response evidence; it did not reverse order and is not a test or replication of Japanese-tit sequence-order sensitivity. No Talk2Nature playback or animal-facing experiment is proposed.

What we can build on

Report the measured receiver behavior and separate it from the theoretical label placed on the behavior. The 2022 source tables descriptively show a large difference between one natural same-source condition and the other tested conditions, but do not choose among Core-Merge, finite-list, concatenation or other mechanisms. A useful next review should assess the 2017 cross-species result alongside the 2016 and 2022 controls, and distinguish linear order, source linkage and hierarchical generalization. Any animal-facing study requires qualified scientific and welfare oversight; these results do not support an AI-generated verbal translation.

OPEN METHODS QUESTION

Independent review of receiver-response evidence

We invite comparative-communication, avian-behavior or animal-cognition researchers to assess what the Japanese-tit experiments support about call order and source linkage; whether the Schlenker and Beckers et al. interpretations make distinguishable predictions; and what the 2017 cross-species and 2025 Carolina-wren/great-tit results do and do not generalize. This is a literature critique, not a Talk2Nature model claim or an animal experiment.

Which 2016 and 2017 contrasts best support an order-sensitive receiver response, and what alternatives remain after the reported component, reversal, novelty and call-type controls?
What does the 2022 one/two-speaker comparison establish about source linkage, and what does it leave open about mechanism or independent replication?
Do Minimal Compositionality and Bird Implicatures (Schlenker et al., 2024) make distinct predictions for the cited experiments? How do Beckers et al. (2024) finite-list and concatenation-based alternatives fit the same evidence, and what does the 2017 cross-species sequence constrain?
How should the 2025 Carolina-wren/great-tit study affect cross-species generalization? Its sequence-versus-D-only comparison did not reverse order, so what does it add about receiver response, and what does it leave open about order or meaning?
What is the smallest public-source check that could change your assessment? If a new animal study is needed, what outcome/control matters and which welfare review would be required before protocol design?

Review status: No independent specialist review has been received. The 2022 descriptive table audit is complete, but its mixed-effects models and adjusted contrasts have not been reproduced. The 2017 publisher page and data-file inventory, the complete 2024 formal supplement, and a direct full-text review of the 2025 article remain unavailable or unreviewed as described in the evidence note.

Read the focused reviewer brief ↗Submit a public methods review ↗

Do not attach recordings, private research material, personal information or sensitive locations. A review is not an endorsement or partnership.

The next reading task

Invite an independent comparative-communication specialist to assess the published interpretations, including the Beckers et al. critique, Schlenker et al. alternatives, the 2017 cross-species sequence and the limits of the 2025 comparator. Before any inferential reanalysis, define a data-to-table mapping, obtain an independently checked analysis plan and freeze the software environment; the current project script is descriptive only. Seek the 2017 Mendeley file inventory and exact file terms only through an authorized route. Any animal study requires approval by the responsible scientific team.

Go to the sources

Bird-call syntax ↗
Suzuki, Wheatcroft & Griesser. Experimental evidence for compositional syntax in bird calls. Nature Communications (2016). Playback experiments in Japanese tits; species-specific evidence.
Bird syntax: systematicity commentary ↗
Phillips & Wilson (2016). Commentary: Experimental evidence for compositional syntax in bird calls. Frontiers in Psychology 7:1171. DOI: 10.3389/fpsyg.2016.01171.
Japanese-tit call combination and source-linkage experiment ↗
Suzuki & Matsumoto (2022), Experimental evidence for core-Merge in the vocal communication system of a wild passerine. A field playback study tested whether Japanese tits' mobbing responses depended on call order and whether two call types came from one or two speaker locations.
Japanese-tit core-Merge study data and analysis code deposit ↗
The 2022 article points to Figshare item 18007046 v1 (DOI 10.6084/m9.figshare.18007046.v1) for two outcome tables and two R scripts. The official API currently lists a record-level CC BY 4.0 license, file sizes and MD5 checksums.
Japanese-tit ordering rule for novel cross-species call sequences ↗
Suzuki, Wheatcroft & Griesser (2017), Wild Birds Use an Ordering Rule to Decode Novel Call Sequences. Japanese tits responded differently to a novel ABC–tää sequence and its reversal; ABC–zi controls tested whether ABC-first alone explained the pattern.
Japanese-tit novel call-sequence experiment data ↗
Mendeley Data v1, DOI 10.17632/r7v96zf5pp.1. The record describes raw playback-experiment data and acoustic analyses and displays CC BY 4.0. The landing page exposes no file rows; an unauthenticated request to the documented public-dataset API returned HTTP 401.
Minimal Compositionality versus Bird Implicatures in Japanese tits ↗
Schlenker et al. (2024), Biological Reviews 99:1278–1297, DOI 10.1111/brv.13068. The theory paper compares Minimal Compositionality with Bird Implicatures for Japanese-tit ABC–D sequences; it reports no new playback experiment.
Read the author-accepted manuscript (HAL v1) ↗
Carolina-wren mobbing sequences and great-tit receiver responses ↗
Coye & Dutour (2025), Mobbing sequences of American wrens elicit mobbing responses in European tits. Publisher-indexed Methods and Results describe a six-treatment playback with 96 wild great tits; direct publisher retrieval returned HTTP 403.
A critique of Core-Merge and language-syntax claims in songbird calls ↗
Beckers et al. (2024), No evidence for language syntax in songbird vocalizations. A published critique of the Japanese-tit call-order and source-linkage interpretation.
Read the full text at PubMed Central ↗

Original editorial note by Talk2Nature contributors. AI-assisted, source-grounded synthesis; not a peer review or an independent replication. Linked works retain their own rights.