Experimental Evidence for Phonemic Contrasts in a Nonhuman Vocal System

The ability to generate new meaning by rearranging combinations of meaningless sounds is a fundamental component of language. Although animal vocalizations often comprise combinations of meaningless acoustic elements, evidence that rearranging such combinations generates functionally distinct meaning is lacking. Here, we provide evidence for this basic ability in calls of the chestnut-crowned babbler (Pomatostomus ruficeps), a highly cooperative bird of the Australian arid zone.

Abstract

The ability to generate new meaning by rearranging combinations of meaningless sounds is a fundamental component of language. Although animal vocalizations often comprise combinations of meaningless acoustic elements, evidence that rearranging such combinations generates functionally distinct meaning is lacking. Here, we provide evidence for this basic ability in calls of the chestnut-crowned babbler (Pomatostomus ruficeps), a highly cooperative bird of the Australian arid zone. Using acoustic analyses, natural observations, and a series of controlled playback experiments, we demonstrate that this species uses the same acoustic elements (A and B) in different arrangements (AB or BAB) to create two functionally distinct vocalizations. Specifically, the addition or omission of a contextually meaningless acoustic element at a single position generates a phoneme-like contrast that is sufficient to distinguish the meaning between the two calls. Our results indicate that the capacity to rearrange meaningless sounds in order to create new signals occurs outside of humans. We suggest that phonemic contrasts represent a rudimentary form of phoneme structure and a potential early step towards the generative phonemic system of human language.

Summary

A major question in language evolution is how its generative power emerged. This power, which allows the communication of limitless thoughts and ideas, is a result of the combinatorial nature of human language: meaningless phonemes can be combined to form meaningful words (phonology), and words can be combined to form higher-order, meaningful structures (syntax). While previous work has indicated the potential for animals to form syntax-like constructions, there exists little convincing evidence for a basic phonemic capacity in animals. Here, we demonstrate, using analyses combined with natural observations and playback experiments, that the cooperatively breeding chestnut-crowned babbler reuses two meaningless acoustic elements to create two functionally distinct vocalizations. This result suggests the basic ability for phoneme structuring occurs outside of humans and provides insights into potential early evolutionary steps preceding the generative phonemic system of human language.

Introduction

The vast lexicons that characterise human languages are the product of physical and cognitive processes that guide the combination of a limited number of meaningless sounds (phonemes) in a variety of ways to generate new meaning [1,2]. In a simple example, the phonemes /k/, /æ/ and /t/ can be rearranged in different ways to create the words cat [kæt], act [ækt] or tack [tæk] [1]. Alternatively, the phoneme /k/ from the word cat can be eliminated to create the word at [æt], with the first position (i.e., presence or absence of the phoneme /k/) representing a phonemic contrast that generates the differentiation in meaning [3]. In all four arrangements, the meaningless phonemes maintain their acoustic identity across words, and this, paired with the arbitrary relationship between phoneme structure and word meaning, results in words with shared phonemes having distinct semantic content [4]. Such phoneme structure is a basic ingredient of word generation in human language, and when combined with the rules governing assemblages of meaningful words (a syntactic layer), provides much of language’s generative power [5–7]. Despite the crucial role that phoneme structure plays in language, little is known about how such a capacity might have evolved [8–11]. Whilst comparative data from animal communication systems can elucidate early forms of language components, data demonstrating the critical rudiments of phoneme structures outside of humans is lacking.

Evidence that animals can employ a basic syntactical layer of language in their communication system has been provided in nonhuman primates. For example, Campbell’s monkeys (Cercopithicus cambelli) produce two predator-specific alarm calls that are each modified in a predictable way into more general disturbance calls upon addition of the same suffix [12,13]. However, because the constituent calls are themselves meaningful (with the suffix carrying an abstract meaning in this case [14]), this, and equivalent findings [15,16], do not exemplify phoneme structure. Several candidates of phoneme-like structures in nonhuman animals have been proposed, but defining features are either lacking or have yet to be demonstrated [8,11,17]. One set of contenders comes from the songs of birds and mammals, in which meaningless elements are combined to create complex, higher-order structures [11,18,19]. However, experiments investigating behavioural responses to element reorganisation within songs are either lacking [18–21] or have not shown that such reorganisation confers a qualitative change in contextual meaning [22–24]. Another set includes calls produced in movement and alarm contexts. For example, parid birds can produce variable vocal sequences of apparently meaningless acoustic elements. However, in these cases, although call elements are commonly repeated or omitted, the required association between sequence structure and qualitative changes in informational content has not been demonstrated [25–29].

Using acoustic analysis, natural observations, and controlled playback experiments we provide evidence for rudimentary phoneme structure in the calls of chestnut-crowned babblers (Pomatostomus ruficeps) (see Materials and Methods), a 50 g, highly social, cooperatively breeding bird [30,31]. Observations over the past 10 years suggest that the repertoire of adult chestnut-crowned babblers consists of at least 15 discrete, context-specific vocalizations, of which three pairs appear to share sound elements, with the reused elements in each case being restricted to a specific pair of calls [32]. Here, we specifically focused on a single pair: a double-element call produced during flight (flight calls, elements F1 and F2) and a triple-element call produced during nestling provisioning [33] (prompt calls, elements P1, P2, and P3) (Fig 1A). Importantly, the constituent elements within these calls appear to be contextually meaningless. For example, none of the elements is used as an individual call in isolation, suggesting that none can function to confer contextual information. Additionally, because none is used in combination with other call types, they cannot clearly operate to modify calls in a predictable way, as would be required of affixes [13]. First, we establish, using acoustic analyses, that the two calls comprise statistically equivalent acoustic elements. Second, we present natural observations showing that the two calls are context-specific, a prerequisite of reliable information transfer in animals. Finally, playbacks of natural, switched-element, and artificial calls in a standardised aviary environment confirm that the call elements are perceptibly equivalent and that element addition/elimination at one position creates a phoneme-like contrast, yielding the functional changes in meaning.

http://journals.plos.org/plosbiology/article/figure/image?size=medium&id=info:doi/10.1371/journal.pbio.1002171.g001

Fig 1. Flight and prompt call structure.

(A) Spectrogram of double-element flight call (sequence F1 F2) and triple-element prompt call (sequence P1 P2 P3), taken from different individuals and groups. (B) Discriminant Function Analysis (DFA) output: function 1 explains 95% of the variance in element structure and primarily describes frequency range; function 2 explains the remaining 5% of variance and describes the contrast between start/end frequency (positive loadings) and frequency range (negative loadings) (Materials and Methods; S1 Text; S1 Table; S2 Table). F1 could not be reliably discriminated from P2 (34% errors: T32 = 1.4, p = 0.2) and nor could F2, P1, and P3 be discriminated from each other (27%–32% errors: F2 versus P1: T32 = 0.7, p = 0.4; F2 versus P3: T32 = 1.4, p = 0.2; P1 versus P3: T20 = 0.2, p = 0.8), but F1/P2 could be easily distinguished from F2/P1/P3 (all 0% errors) (T32–44 = 14.1–22.9; all p values < 0.001; S3 Table). Accordingly, flight calls and prompt calls follow AB and BAB construction, respectively.

Results

Acoustic Analysis

Acoustic analyses were conducted to test whether prompt and flight calls are composed of statistically indistinguishable acoustic elements. To avoid problems of pseudo-replication arising from using calls of genetic relatives within groups [34], we analysed a single flight call and a single prompt call per group recorded (n = 23 flight, 11 prompt calls). Five parameters were extracted from the fundamental frequency of the resulting 79 elements: start and end frequency, frequency range, time to peak frequency, and element duration (S1 Text, S1 Table, and S2 Table). A Discriminant Function Analysis (DFA) demonstrated that the five elements across the two calls comprised just two independent acoustic structures (Fig 1B). Mahalanobis distances generated from the DFA revealed that F1 and P2 could not be reliably distinguished and neither could F2, P1, and P3 (all p values > 0.2), but that F1 and P2 could be distinguished easily from F2, P1, and P3 (all p values < 0.001) (Fig 1B and S3 Table). Thus, the two calls appear to comprise the same two distinct elements, with flight and prompt calls displaying AB and BAB patterns, respectively.

Natural Observations

Natural observations were conducted to quantify the context in which flight and prompt calls are produced. Natural flights were accompanied by flight calls in 274 of 450 observations (61%; n = 6 groups, 1 h/group), with all flights being short, low, and easily quantified. Similarly, hand-held releases following capture induced flight calls in 58 of 90 occasions (64%, n = 24 groups). No prompt calls were recorded in either set of observations, and flights/releases lacking flight calls were either silent or associated with alarm calls in response to observer presence. Finally, recordings from within nests in conjunction with automated nest entry-exit recorders revealed that 62% flights to/from nests were accompanied by flight calls (n = 140 visits, 7 groups) but rarely prompt calls (0.08% of nest visits), while 70% of nestling provisioning events were associated with prompt calls (n = 140 visits, 7 groups) and rarely flight calls (0.03% of nest visits). Additionally, in 97% of nest visits in which both flight and prompt calls were recorded, individuals used only flight calls travelling to/from the nest and only prompt calls within nests (n = 60 visits, 7 groups). Thus, flight and prompt calls are highly context-specific, with the former maintaining group cohesion during movement [35] and the latter increasing the efficiency of food transfer to offspring by stimulating begging [33].

Playback Experiments

To verify experimentally that flight and prompt calls are context specific and are generated from rearrangement of the same acoustic elements, we performed playback experiments on 16 birds captured from 7 groups during periods of breeding. Each of the 16 birds received six playback trial-sets presented in a randomised order. Behavioural responses to two natural, two switched-element and two artificial calls were recorded in aviary compartments (2 x 2.5 x 2 m l x b x h) containing natural perches, foraging substrate, a view to the outside, and a recently used babbler nest (30×45 cm dome-shape, 6 cm diameter entrance) (Fig 2). The playback speaker was positioned out of view in a neighbouring compartment; birds had to look perpendicular to the speaker to look outside the aviary and in the opposite direction to look at the nest (S2 Text). Given our natural observations, we predicted flight calls would elicit increased observations to the outside and increased movement in anticipation of an incoming bird, while prompt calls would provoke greater nest attentiveness. Combined, these three behaviours comprised 61% of the activity budget in each trial (SD [standard deviation] = 23%; correlation coefficients among these behaviours ranged from +0.1 to -0.3, indicating that time spent in one activity did not preclude time available for another).

http://journals.plos.org/plosbiology/article/figure/image?size=medium&id=info:doi/10.1371/journal.pbio.1002171.g002

Fig 2. Schematic of aviary setup.

The aviary consisted of six compartments: the back comprised metal meshing (1 cm2), allowing the birds an outside view; the two sides were made of aluminium; and the front was specially designed perspex, allowing a one-way view from outside to inside. Occupied compartments contained natural perches, foraging substrate, a feeding station, babbler nest, and sleeping box, while unoccupied compartments contained the playback apparatus. Babbler nests are large (~45 x 30 cm), dome-shaped, with 6 cm diameter entrance hole, and robust. Babblers spent most of their time at mid-height; in all cases, relative to the speaker, birds had to look behind and up to look at the nest. Single birds used compartment 3 (n = 2), pairs of birds used compartments 1 and 3 (n = 1 pair) and trios used compartments 1, 3, and 6 (n = 4 trios) (S2 Text).

Compared with natural prompt calls, natural flight call playbacks were associated with a 49% increase in the proportion of time spent looking outside (Generalized Linear Mixed Model [GLMM]: χ21 = 11.8, p < 0.001) and a 36% increase in time spent hopping/flying between perches (χ21 = 6.5, p = 0.02). By contrast, during natural flight call playbacks, individuals spent 81% less time looking at the nest (2% of monitoring time) than during prompt call playbacks (15% of time) (χ21 = 11.6, p < 0.001) (Fig 3). Together, these results confirm the two calls are distinct and encode perceptible, context-specific information.

http://journals.plos.org/plosbiology/article/figure/image?size=medium&id=info:doi/10.1371/journal.pbio.1002171.g003

Fig 3. Responses to natural playbacks.

Proportion of time spent engaged in three behaviours of functional relevance differed significantly during the playbacks of the two call types (see text). Figure shows back-transformed predicted means (± standard error [SE]) generated from three Generalized Linear Mixed Models (GLMM), in which the time engaged in each of the three activities (looking out of the aviary, in movement [hopping/flying], and looking at the nest) were fitted as three independent response terms. In each case, response terms were fitted to a binomial error structure with logit link function, time spent in camera view was fitted as the binomial denominator, call type (natural flight versus natural prompt) was fitted as a two-level factor, and individual identity nested within group identity were fitted as random terms.

To test whether unmeasured acoustic variation dissociates the two calls [15], we played back switched-element versions of both calls to all 16 birds by generating flight calls from prompt elements P2 P3 and prompt calls from elements P1 F1 F2. The proportion of time birds spent engaged in the three behaviours of functional relevance were statistically equivalent between natural and switched-element flight calls (GLMM: all p values > 0.6; Fig 4A) as well as between natural and switched-element prompt calls (all p values > 0.3; Fig 4B). Additionally, there were no significant interactions between call type (flight versus prompt) and whether or not calls were natural or switched-element on behavioural responses (GLMM: all p values > 0.4). The absence of such interactions generated differences in behavioural responses to switched-element flight versus switched-element prompt calls of similar magnitude to those found in comparisons of natural calls (see Fig 4C versus Fig 3). Compared with switched-element prompt calls, switched-element flight calls were associated with 33% more time looking out, 33% more time in-movement, and 80% less time looking at the nest. Accordingly, it is improbable that any unmeasured acoustic differences between the elements of flight and prompt calls are responsible for the distinct responses, reinforcing our acoustic analyses that the calls comprise the same sound elements.

thumbnail

Fig 4. Responses to calls with and without reciprocal element exchange.

Behavioural responses of functional relevance remained similar between (A) natural flight calls and switched-element flight calls comprising prompt call elements (natural versus switched-element comparisons: look out: χ21 = 0.02, p = 0.9; hop/fly: χ2 1 = 0.2, p = 0.6; look nest: χ21 = 0.03, p = 0.9) and (B) natural prompt calls and switched-element prompt calls using the two flight call elements (natural versus switched-element comparisons: look out: χ21 = 0.2, p = 0.6; hop/fly: χ21 = 1.3, p = 0.3; look nest: χ21 = 0.01, p = 0.9). (C) Behavioural responses to switched-element flight and switched-element prompt calls differed significantly or showed a non-significant tendency to do so (switched-element flight versus switched-element prompt call comparisons: look out: χ21 = 5.7, p = 0.02; hop/fly: χ21 = 3.2, p = 0.09; look nest: χ21 = 10.0, p = 0.002). Analyses were conducted as above (Fig 3) except that in (A) and (B) (which are shown separately for clarity) stimulus type (natural versus switched-element) and its interaction with call type were added as additional fixed effects, while in (C), natural flight and prompt calls were replaced with switched-element ones. All figures show back-transformed predicted means (±SE).

The results above suggest that the meaning-differentiating element between the two calls is P1. Before a phonemic-like system can be supported, two other interpretations require testing. First, element P1 might, by itself, be responsible for generating the contextual information carried by the prompt call, in which case, our results could be more akin to a syntactic, rather than phonemic, communicative system [12,13]. Second, the differences in response to flight calls versus prompt calls might arise from their differences in element number [36]. In this case, our results could represent stimulus intensity effects (triple-element prompt versus double-element flight call) or priming effects [12] (any acoustic element preceding a flight call results in a prompt-type response). To test these alternative interpretations, we presented two artificial stimuli to the 16 birds: element P1 alone and CAB, with the latter representing call elements P2 P3 (i.e., AB) preceded by an element (C) from chatter calls, a common call naturally repeated in mixed-element bouts and associated with excitement [32].

These two artificial stimuli elicited similar behavioural responses (all p values > 0.2; Fig 5A). First, they both generated relatively high look out and movement responses. One explanation lies with the fact that each is unnatural: impossible vocal scenarios have been shown to increase attentiveness behaviour in other contexts [37,38]. In support, separate analysis of the proportion of time spent looking around the aviary showed that general attentiveness behaviour during natural flight playbacks (mean ± SE = 16% ± 4%) was 36%, 47%, and 48% lower than during playbacks of CAB, P1, and natural prompts, respectively (GLMM: χ23 = 10.6, p = 0.01). Second, and more crucially, neither the P1- nor the CAB-stimulus elicited a hint of an elevated response in nest-attentiveness (Fig 5B). Like the flight call, P1 element and CAB playbacks were both associated with ca. 80% reductions in nest-attentiveness behaviour over natural prompt calls (Fig 5B). That neither the P1 element alone nor CAB elicits any increase in nest-attentiveness confirms that (a) P1 does not carry any nest-associated information in isolation and (b) differential nest-attentiveness responses to flight and prompt calls are not derived from either stimulus intensity or priming effects. Thus, it is the presence or absence of element P1 from the P2 P3 (or F1 F2) call sequence that appears integral to generating the qualitatively distinct meaning carried by the two calls.

http://journals.plos.org/plosbiology/article/figure/image?size=medium&id=info:doi/10.1371/journal.pbio.1002171.g005

Fig 5. Behavioural responses to artificial calls.

(A) During P1 element and CAB playbacks, individuals spent comparable proportions of time looking out (χ21 = 0.2, p = 0.6), in movement (χ21 = 1.8; p = 0.2), and looking at the nest (χ21 = 0.5; p = 0.5). (B) Neither P1 nor CAB playbacks provoked an increase in nest-attentiveness over natural flight calls, leading to proportions of time spent looking at the nest during these trials being substantially lower than those generated during natural prompt calls (χ23 = 25.4; p < 0.001). Figures show back-transformed predicted means (±SE) generated from GLMM analyses as outlined in Fig 3 (A) or in which the two-level factor was replaced with a four-level factor (B). 

Discussion

Phoneme structure represents a critical component of the vast lexicons in human languages, but a lack of suitably comparable evidence in animals has hindered our understanding of candidate selection pressures on, and early forms of, phoneme structure. Two related hypotheses have been proposed to explain the emergence of phonemic systems; both advocate a role of selection acting on increasing the capacity of vocal communication beyond that currently possible under an existing vocal repertoire. The here-named “enhanced-perception hypothesis” proposes that stringing together existing sounds in new ways reduces perception errors over the generation of new, but similar, sounds [39–41]. By contrast, the “vocal-constraints hypothesis” proposes that when the generation of new sounds is constrained [42], reusing pre-existing sounds in new combinatorial forms can provide an alternative solution to increasing communicative output [15,16]. Testing the predictions arising from these hypotheses represents a major challenge because human languages are generally too derived to address the pressures selecting for their emergence. Additionally, testing whether animals make perceptual mistakes for sounds that do not exist or are vocally constrained will be rarely feasible. A necessary first step in elucidating the pressures selecting for, and early forms of, phonemic structure is to address whether animals possess the capacity for generating functionally distinct vocalizations by rearranging contextually meaningless elements, and how such rearrangements are manifest.

Here, using acoustic analyses, natural observations, and playback experiments, we reveal that chestnut-crowned babblers use two acoustic elements (A and B) in different arrangements to create two functionally distinct vocalizations: flight calls (F1 F2, or AB) and prompt calls (P1 P2 P3, or BAB). The meaning differentiation between the two calls is not a result of the different number of elements or priming effects, but specifically the presence or absence of P1 (element B) at the head of the same call sequence. The fact that element P1 is both contextually meaningless on its own and meaning differentiating when used in combination with elements P2(F1) and P3(F2) signifies a phoneme-like contrast, with element B used in this position likely representing a phoneme-equivalent. To our knowledge, this is the first demonstration that animals have the basic capacity to use phoneme-like contrasts to derive qualitatively new meaning, a basic component of phoneme structuring. However, whether or not our results can also be interpreted as providing evidence for more advanced forms of phoneme structuring in an animal depends on two critical features.

First, in human languages, phoneme structure has potentially boundless generative power: the sum of derivable information is substantially greater than the number of its phonemic parts [1]. In contrast, the babbler vocal system that we describe is strictly bounded in its generative nature (i.e., two elements generate only two distinct calls). Part of the difference in human versus any nonhuman phonemic system will inevitably arise from vast differences in cognitive capacity [9]. Notwithstanding, cognitive capacity alone does not appear to be sufficient to explain differences in phonemic complexity and boundedness. For example, the sign language of the Al Sayyid Bedouin, an emerging language shared by deaf and hearing people of a small Israeli village, has been shown to have a fully functional and productive syntactic layer, but is so far characterized by only one phonological form [43,44]. Thus, when a phonemic layer emerges, even in human language, it appears initially to be finite and strongly bounded. This evidence suggests that the use of phonemic structure in communication should not be defined a priori by its complexity or boundedness, for it is likely that all phonemic systems evolve from simple beginnings like the one we describe here.

Second, the level of phonemic complexity used by babblers depends on the number of phoneme-equivalent entities in use. For example, whilst babblers generate a phonemic contrast by inserting the phoneme-like entity P1 before P2(F1) and P3(F2), whether or not P2 and F1 or P3 and F2 also represent phoneme-equivalent entities in the linguistic sense is equivocal. Unlike combinatoriality based on affixation rules or the generation of idioms, in which constituent parts have meaning [12,16], definitively testing whether all sound elements within call sequences of animals are contextually meaningless, and yet individually perceptible and meaning-differentiating, will be a major challenge. This is because any sound uttered by a conspecific can lead to a behavioural response irrespective of any perception of contextual meaning [38], and their limited vocal repertoires preclude investigation of whether distinct functional meaning is derived from the same meaningless elements in multiple different arrangements. A key component in discerning whether F1/P2 and F2/P3 are also phonemic depends on whether they represent a compound of two discrete elements, perceptible independently (i.e., A and B), or a holistic unit (i.e., AB). That the B element is phonemic in position P1 hints that AB is reducible, and hence F1/P2 and F2/P3 are probably also phoneme-like. However, this is an untested hypothesis at this stage, and we do not wish to speculate on whether chestnut-crowned babblers use more advanced forms of phoneme structure, beyond the identified use of a simple contrast, as part of their communication system.

Either way, we propose that the bounded use of phoneme-like contrasts in the vocal system of chestnut-crowned babblers represents a simple precursor of phoneme structuring that can elucidate how early forms of phonemic systems might emerge. For example, our results lead to the hypothesis that the addition or elimination of elements, i.e., basic phonemic contrasts (e.g., /kæt/ versus /æt/), might represent a simpler evolutionary step than complete element rearrangement (e.g. /kæt/ versus /tæk/), due to its reduced structural complexity. However, generating distinct contextual meaning through the former rather than latter process is likley to be more prone to perception errors, because it results in higher acoustic similarity. That babblers have opted for the more error-prone means of generating functionally distinct vocalizations, and done so by adding or eliminating a common element, is more supportive of a vocal-constraints hypothesis [15,16] than an enhanced-perception hypothesis [39–41]. Limiting the use of phonemic contrasts to short-range calls used in low-urgency, social contexts might be one way of reducing perception errors and mitigating associated costs when vocal constraints are operating.

In conclusion, the salient message here is that the basic capacity to generate qualitatively new meaning from rearranging contextually meaningless elements appears to exist outside of humans. One explanation is that for vocally constrained, highly social species, such as chestnut-crowned babblers, evolving new meaning by rearranging existing sounds offers a faster route to increasing communicative output than evolving new sounds. We hypothesise that reusing acoustic elements has facilitated the emergence of phoneme-like contrasts, which potentially drove sensitivity to phoneme structure or “phonemic awareness” in receivers [45,46]. The capacity to recognise vocalizations as sound constructs composed of smaller, meaningless elements, instead of a holistic unit, may have been the first step in the emergence of the elaborate phonemic systems seen in human languages. Further experiments are now required to determine exactly how babblers compute and perceive the elements from the two calls. More generally, further evidence for the use and manifestation of phonemic systems in animals is required; we propose that such systems will be most operant in the short-range communication of vocally constrained, social animals.

References

  1. Hockett CF (1960) The Origin of Speech. Scientific American 203: 88–111. pmid:13683472
  2. Hauser MD (1997) The Evolution of Language. Cambridge: MIT Press.
  3. de Boer B, Sandler W, Kirby S (2012) New perspectives on duality of patterning: Introduction to the special issue. Language and Cognition 4: 251–259. pmid:24587833
  4. Hurford J (2007) The origins of meaning. Oxford, New York: Oxford Univiversity Press.
  5. Chomsky N (1981) Knowledge of Language: Its Elements and Origins. Philosophical Transactions of The Royal Society B: Biological Sciences 295: 223–234.
  6. Nowak MA, Plotkin JB, Jansen VAA (2000) The evolution of syntactic communication. Nature 404: 495–498. pmid:10761917
  7. Hurford J (2011) Linguistics from an Evolutionary Point of View. In: Kempson R, Fernando M, Ashe N, editors. Handbook of the Philosophy of Science: Linguistics. Amsterdam: Elsevier Science.
  8. Marler P (1977) The structure of animal communication sounds. Recognition of complex acoustic signals: report of Dahlem workshop. Berlin: Abakon Verlagsgesellschaft.
  9. Hauser MD, Chomsky N, Fitch WT (2002) The Faculty of Language: What Is It, Who Has It, and How Did It Evolve? Science 298: 1569–1579. pmid:12446899
  10. Yip MJ (2006) The search for phonology in other species. Trends in Cognitive Sciences 10: 442–446. pmid:16931109
  11. Berwick RC, Beckers GJL, Okanoya K, Bolhuis JJ (2012) A bird’s eye view of human language evolution. Frontiers in Evolutionary Neuroscience 4: 5. pmid:22518103
  12. Ouattara K, Lemasson A, Zuberbühler K (2009) Campbell’s Monkeys Use Affixation to Alter Call Meaning. PLoS ONE 4: e7808. pmid:19915663
  13. Ouattara K, Lemasson A, Zuberbühler K (2009) Campbell’s monkeys concatenate vocalizations into context-specific call sequences. Proceedings of the National Academy of Sciences 106: 22026–22031. pmid:20007377
  14. Collier K, Bickel B, van Schaik CP, Manser MB, Townsend SW (2014) Language evolution: syntax before phonology? Proceedings of the Royal Society B: Biological Sciences 281: 20140263. pmid:24943364
  15. Arnold K, Zuberbühler K (2006) Language evolution: Semantic combinations in primate calls. Nature 441: 303. pmid:16710411
  16. Arnold K, Zuberbühler K (2008) Meaningful call combinations in a non-human primate. Current Biology 18: R202–R203. pmid:18334192
  17. Catchpole CK, Slater TLB (1995) Bird Song: Themes and variations. New York: Cambridge University Press.
  18. Payne RS, McVay S (1971) Songs of Humpback Whales. Science 173: 585–597. pmid:17833100
  19. Geissmann T (2002) Duet-splitting and the evolution of gibbon songs. Biological Reviews 77: 57–76. pmid:11911374
  20. Leger DW (2005) First Documentation of Combinatorial Song Syntax in a Suboscine Passerine Species. The Condor 107: 765–774.
  21. Trillo PA, Vehrencamp SL (2005) Song types and their structural features are associated with specific contexts in the banded wren. Animal Behaviour 70: 921–935. pmid:17173097
  22. Holland J, Dabelsteen T, Paris AL (2000) Coding in the song of the wren: importance of rhythmicity, syntax and element structure. Animal Behaviour 60: 463–470. pmid:11032649
  23. Briefer EF, Rybak F, Aubin T (2013) Does true syntax or simple auditory object support the role of skylark song dialect? Animal Behaviour 86: 1131–1137.
  24. Lachlan RF, Nowicki S (2015) Context-dependent categorical perception in a songbird. Proceedings of the National Academy of Sciences 112: 1892–1897. pmid:25561538
  25. Hailman JP (1989) The Organization of Major Vocalizations in the Paridae. The Wilson Bulletin 101: 305–343.
  26. Ficken MS, Hailman ED, Hailman JP (1994) The Chick-a-Dee Call System of the Mexican Chickadee. The Condor 96: 70–82.
  27. Freeberg TM, Lucas JR (2002) Receivers respond differently to chick-a-dee calls varying in note composition in Carolina chickadees, Poecile carolinensis. Animal Behaviour 63: 837–845.
  28. Templeton CN, Greene E, Davis K (2005) Allometry of Alarm Calls: Black-Capped Chickadees Encode Information About Predator Size. Science 308: 1934–1937. pmid:15976305
  29. Suzuki TN (2013) Communication about predator type by a bird using discrete, graded and combinatorial variation in alarm calls. Animal Behaviour 87: 59–65.
  30. Browning LE, Patrick SC, Rollins LA, Griffith SC, Russell AF (2012) Kin selection, not group augmentation, predicts helping in an obligate cooperatively breeding bird. Proceedings of the Royal Society B: Biological Sciences 279: 3861–3869. pmid:22787025
  31. Rollins LA, Browning LE, Holleley CE, Savage JL, Russell AF, et al. (2012) Building genetic networks using relatedness information: a novel approach for the estimation of dispersal and characterization of group structure in social animals. Molecular Ecology 21: 1727–1740. pmid:22335253
  32. Crane JMS (2014) Vocal communication in the cooperatively breeding chestnut-crowned babbler [PhD thesis]: University of Sheffield, PhD.
  33. Young CM, Browning LE, Savage JL, Griffith SC, Russell AF (2013) No evidence for deception over allocation to brood care in a cooperative bird. Behavioral Ecology 24: 70–81.
  34. Sharp SP, McGowan A, Wood MJ, Hatchwell BJ (2005) Learned kin recognition cues in a social bird. Nature 434: 1127–1130. pmid:15858573
  35. King BR (1980) Social-Organization and Behavior of the Grey-Crowned Babbler Pomatostomus-Temporalis. Emu 80: 59–76.
  36. Slocombe KE, Kaller T, Call J, Zuberbuhler K (2010) Chimpanzees Extract Social Information from Agonistic Screams. PLoS ONE 5: e11473. pmid:20644722
  37. Cheney DL, Seyfarth RM, Silk JB (1995) The responses of female baboons (Papio cynocephalus ursinus) to anomalous social interactions: evidence for causal reasoning? Journal of Comparative Psychology 109: 134–141. pmid:7758289
  38. Townsend SW, Allen C, Manser MB (2012) A simple test of vocal individual recognition in wild meerkats. Biology Letters 8: 179–182. pmid:21992821
  39. Nowak MA, Krakauer DC, Dress A (1999) An error limit for the evolution of language. Proceedings of the Royal Society B: Biological Sciences 266: 2131–2136. pmid:10902547
  40. Nowak MA, Krakauer DC (1999) The evolution of language. Proceedings of the National Academy of Sciences 96: 8028–8033. pmid:10393942
  41. Pinker S, Jackendoff R (2005) The faculty of language: what’s special about it? Cognition 95: 201–236. pmid:15694646
  42. Janik VM, Slater PJ (2000) The different roles of social learning in vocal communication. Animal Behaviour 60: 1–11. pmid:10924198
  43. Sandler W, Aronoff M, Meir I, Padden C (2011) The gradual emergence of phonological form in a new language. Natural Language & Linguistic Theory 29: 503–543.
  44. Sandler W, Aronoff M, Padden C, Meir I (2014) Language emergence: Al-Sayyid Bedouin Sign Language. In: Enfield NJ, Kockelman P, Sidnell J, editors. The Cambridge Handbook of Linguistic Anthropology. Cambridge: Cambridge University Press. pp. 246–278.
  45. Anthony JL, Francis DJ (2005) Development of Phonological Awareness. Current Directions in Psychological Science 14: 255–259.
  46. Pepperberg IM (2007) Grey parrots do not always ‘parrot’: the roles of imitation and phonological awareness in the creation of new labels from existing vocalizations. Language Sciences 29: 1–13.
  47. Browning LE, Young CM, Savage JL, Russell DJF, Barclay H, et al. (2012) Carer provisioning rules in an obligate cooperative breeder: Prey type, size and delivery rate. Behavioral Ecology and Sociobiology 66: 1639–1649.
  48. Sorato E, Gullett P-R, Creasey MJS, Griffith SC, Russell AF (2015) Plastic territoriality in group-living chestnut-crowned babblers: roles of resource value, holding potential and predation risk. Animal Behaviour 101: 155–168.
  49. Nomano FY, Browning LE, Savage JL, Rollins LA, Griffith SC, et al. Unrelated helpers neither signal contributions nor suffer retribution in chestnut-crowed babblers. (2015) Behavioral Ecology.
  50. Engesser S (2015). Data from: Experimental evidence for phonemic contrasts in a non-human vocal system. Dryad Digital Repository. Openly available via http://dx.doi.org/10.5061/dryad.082v2.
  51. Schielzeth H (2010) Simple means to improve the interpretability of regression coefficients. Meth Ecol Evol 1: 103–113.

Leave a Reply

Experimental Evidence for Phonemic Contrasts in a Nonhuman Vocal System

The ability to generate new meaning by rearranging combinations of meaningless sounds is a fundamental component of language. Although animal vocalizations often comprise combinations of meaningless acoustic elements, evidence that rearranging such combinations generates functionally distinct meaning is lacking. Here, we provide evidence for this basic ability in calls of the chestnut-crowned babbler (Pomatostomus ruficeps), a highly cooperative bird of the Australian arid zone.

Scroll to top