The study of phonetic universals is important to the science of speech for several interrelated reasons involving the nature of communication, the mental representation of grammar, language change, and the mechanics of speaking and signing. Phonetic patterns that are common to many or all languages are potentially revealing about the cognitive, physiological, physical, and social forces that determine the form of language. In the field of linguistics, universals and tendencies are important both to linguistic typology and to generative linguistics, and it may not be obvious how universals are interpreted differently in these different linguistic traditions. Within generative linguistics, the search for universals is part of the process of establishing the nature of the the human faculty for language, with the idea that the phonetic realization of language is meaningfully constrained by limitations on the mental representation of phonetics. For example, if humans are hardwired to expect spoken languages to be organized into consonant-vowel sequences, then the universality or near- universality of CV syllables in spoken languages is taken to be confirmation of this. Within phonetics and typology, the term “universals” is often used in connection with strong tendencies. These strong tendencies reflect the forces that shape language. For example, if salient modulations in the transition from consonant to vowel are useful for human auditory systems trying to decode speech, then CV sequences are expected to be prevalent (but not necessarily strictly universal) in spoken languages.

Maddieson (1997) uses the term mechanistic universals for phonetic tendencies that are attributed to the speech and hearing mechanisms and the laws of physics. In contrast, ecological universals are attributed to the functional demands of communication. Maddieson explains the fact that more posterior stops such as /k/ typically have longer voice onset time than anterior stops such as /p/ by appealing to aerodynamics and the mechanics of the jaw lowering motion. This is a mechanistic universal. On the other hand, the fact that spoken languages make use of stops at different places of articulation and intersperse them with vowels are attributed to two of Maddieson’s ecological factors, namely that in order to encode meaning, languages must contain recognizable parts that can be distinguished from each other, and that language must also produced in a continuous stream. In this example, the ecological factors are expected to apply (in some form) to both signed and spoken languages, but the specific mechanistic factors about voice onset time are due to mechanisms found only in the spoken modality.

Hyman (2008) distinguishes descriptive universals, which are often readily identifiable from the phonetic signal, from analytic universals, which depend on a particular theory of the mental representation of phonetics and/or phonology. To illustrate this, Hyman considers the claim that all spoken languages have voiceless stops: If this is treated as a descriptive universal, it is falsified by a language whose stop consonants are consistently produced with voicing, but if it is treated as an analytic universal, such a language may be interpreted as having underlying voiceless stops that consistently undergo a voicing process. Imagining this claim as an analytic universal provides an opportunity to appreciate the tension between two approaches to universals: Is the postulation of underlying voiceless stops a trick to preserve a universal that is not descriptively true? Or is it taking the mental representation seriously in a way that cannot be addressed solely in terms of the phonetic signal? Kiparsky (2018) argues that it only makes sense to think of universals in terms of a formal model of the mental representation of phonetics and phonology, and that there really are no descriptive universals. This is a source of tension within the field: If you are committed to developing a formal model of the mental representation of language, you might be frustrated by more descriptive approaches that you see as superficial and failing to take into account the reality that language is mentally represented. If you are committed to exploring external forces on language, you might be frustrated by more formal approaches that you see as mistaking mechanistic and ecological patterns for cognitive ones. Evans and Levinson (2009) and its 23 commentaries provide a good example of the ongoing debate over universals, including much debate over what the debate even is.

A particularly influential articulation of the primacy of analytic universals is found on pages 4-5 of Chomsky and Halle (1968): Linguistic theory is a theory of the mental representation of language. The nature of this representation limits the set of possible languages and limits the search space for language-learning children. Chomsky and Halle distinguish formal universals, which determine the structure of grammars and rules, from substantive universals, which define the elements that grammars refer to, such as syntactic categories and the phonetic features that define possible speech sounds. If true, the idea that there is a closed set of phonetic representations would be expected to yield some important phonetic universals.

Based on an analysis of the phoneme inventories in the UPSID database (Maddieson and Precoda, 1990), Hyman (2008, 102) concludes that “[a]ll languages have stops and coronals, and all vowel systems exploit vowel height. In the case of consonants, another feature is required (e.g., voice, nasality), as well as a second place of articulation besides coronal.” This is universal to the 451 spoken language inventories in the UPSID database (representing about 7% of the world’s spoken languages), and therefore it must be a very strong tendency even if it doesn’t turn out to be universal to every spoken language. Hyman emphasizes that what makes the observation about stops interesting is that there are no corresponding universals involving other manners of articulation like fricatives and nasals. Some additional things Hyman shows are present in all UPSID inventories are (1) a back vowel, (2) a front vowel or palatal glide, and (3) an unrounded vowel (actually 2 is the observed minimum).

Maddieson (1997) describes many of the strong phonetic tendencies in spoken languages that are often discussed as universals, including (1) higher vowels have higher fundamental frequency and shorter duration than lower vowels, (2) higher vowels have a greater tendency to devoice than lower vowels (see also Jaeger 1978), (3) fundamental frequency is higher after a voiceless consonant than after a voiced one, (4) vowels are longer before voiced consonants and shorter in closed syllables, (5) more anterior stops have longer closure intervals and shorter voice onset times than more posterior stops (see also Cho and Ladefoged 1999), (6) speech intervals are longer before prosodic boundaries, and (7) fundamental frequency and overall acoustic amplitude decline through the course of an utterance.

Kawasaki (1982) and Kawasaki-Fukumori (1992) argue that the prevalence of sequences of segments crosslinguistically is determined by the degree of acoustic modulation between those segments. Some universals of gesture timing have been proposed for segments with multiple gestures: Plauché et al. (1998) report that glottalized sonorants are usually produced with preglottalization when prevocalic and with postglottalization when postvocalic, and Gick et al. (2006) report that postvocalic liquids typically have dorsal constrictions, and the timing of liquid gestures typically follows a pattern where more posterior constrictions are closer to syllable nuclei while anterior constrictions are closer to syllable edges.

Maddieson (1978) proposes several universals involving tone, including (1) the maximum number of contrasting level tones is five, (2) more tone levels require a larger pitch range than fewer tone levels, and (3) contour tones imply level tones and complex contours imply simple contours. See also Hyman and Schuh (1974) and Cahill (2007) for proposed universals of tone. Vogel (2009) offers ten analytic universals for prosodic structures, and Booij (1983) discusses some mechanistic and/or ecological explanations for prosodic universals.

2.1 Mechanistic explanations

Ohala (1983) describes mechanistic explanations for a wide variety of sound patterns, including many of those in Maddieson’s list above. Many of these are rooted in the aerodynamic voicing constraint, which states that vocal fold vibration requires airflow through the glottis, which in turn requires a positive difference between subglottal and intraoral pressure, something which is compromised by the production of obstruent consonants (see Ohala 2011). This is a classic mechanistic explanation: It is rooted in the laws of physics and how they play out inside the human vocal tract during speech production. This means that the effect should be independently observable from typological data and from biomechanical simulations. To explore this, Westbury and Keating (1986) made an electrical model of the aerodynamics of stops produced in word-initial, word-medial, and word-final positions and showed that word-medial position favors voicing and the other positions mostly favor voicelessness. Keating (1984a) similarly showed that voicing is favored at more anterior places of articulation and that this can be attributed to the greater surface area of the vocal tract walls behind more anterior constrictions.

Despite the great interest in mechanistic explanations for linguistics universals, it has been observed routinely that proposed mechanical effects are not universally reflected among the world’s languages, or even just among the world’s spoken languages. Some languages match the magnitudes of the predicted effects closely, some lack them entirely, and some appear to exaggerate them. After reviewing three such cases, Keating (1984b, 45) suggests "that we consider all phonetic processes, even the most low-level, to be phonologized (or grammaticized) in the sense that they are cognitively represented, under explicit control by the speaker, and once-removed from – that is, not automatic consequences of – the physical speaking machine."

Investigating phonologized patterns, Solé (2007) reviews several cases of phonetic/phonological effects (such as anticipatory nasalization) that appear to reveal the universal/mechanical version of the effect in one language and are clearly phonologized in another, and Chodroff et al. (2019) show that languages are consistent across laryngeal categories in how they phonologize the relationship between voice onset time and place of articulation. Ladefoged (2002) argues that understanding phonetic universals will require a better understanding of the neurological basis of speech gestures and how they are executed.

Greenberg (1966b) described one of the ways for common phonological patterns to be "once-removed" from mechanistic factors: he postulated that many seemingly unrelated observations about nasal vowels can be explained in terms of the observation that they arise mostly or exclusively from a sound change such as VN > ṼN > Ṽ. This thinking has formed the basis for a framework in which the nature of sound systems are understood in terms of common changes. Javkin (1978) made a taxonomy of articulatorily-based, acoustically-based, and perceptually-based sound changes, subsequently elaborated to include aerodynamically-based changes by Jaeger (1978). Blevins (2004) incorporated these categories of sound change and an additional mechanism (Chance, in which the mental representation changes but the phonetic output does not), in the Evolutionary Phonology framework. Moreton (2008) groups all of the factors related to production and perception of speech as "channel bias", as opposed to "analytic bias", which reflects learning biases and constraints on possible mental representations (i.e., Chomsky and Halle’s formal and substantive universals).

2.2 Ecological explanations

Returning to Maddieson’s ecological universals mentioned above, the need to produce recognizable distinguishable parts in a continuous stream is often viewed as conflict (or compromise) between the need for perceptual distinctiveness and the need for articulatory efficiency (see Lindblom et al. 1983 for elaboration of these ideas). Liljencrants and Lindblom (1972) modeled vowel categories as charged particles constrained to a realistic articulatory space. Through repulsion (i.e., the demand for perceptual distinctiveness) the vowels distribute themselves in the available space (i.e., as constrained by the demand for articulatory efficiency), resulting in configurations that are similar to common spoken language vowel inventories. Given a number of vowel phonemes (such as 5) the model provides a reasonable approximation of the phonetic qualities of the vowels (such as [i ɛ æ/a ɑ u]).

An alternative to the strictly ecological explanation is that the need for dispersedness is encoded in the grammar (Flemming, 2002, 2004), or that it is a consequence of phonological representations: Hall (2011) argues that dispersedness of phoneme inventories arises as a result of a bias toward constrastive features in phonological representations. Another instantiation of this representational approach is that phoneme inventories are limited to those that can be distinguished by a small set of universal distinctive features (Chomsky and Halle, 1968; Duanmu, 2016).

Lindblom and Maddieson (1988) showed that it is harder to predict consonant inventories from ecological constraints than to predict vowel inventories, but there are some common patterns, such as the fact that obstruents typically comprise about 70% of the consonant phonemes in a spoken language inventory, which they attribute to the nature of the regions of phonetic space where obstruents and sonorant consonants exist. Nevertheless, it is possible to identify a core set of consonants that are used in very many spoken languages. Bybee and Easterday (2022) observe that many non-basic consonants commonly arise by sound change from basic consonants, but that many of the basic consonants (/p t k b d g m n ŋ s l/) are rarely created by sound change, and they suggest that these consonants may have been present since the earliest spoken language.

Although both signed and spoken languages are considered to have phonetics and phonology, many proposed universals and strong tendencies that have been reported in the literature are limited to spoken languages. Often this limitation is implicit. Any universals that apply to both signed and spoken languages have the potential to separate cognitive or ecological constraints from modality-specific factors, and so the existence of signed and spoken modalities provide an opportunity to distinguish modality-specific effects from modality-independent cognitive constraints (see also Sandler 2017).

In chapter 16 of the book Sign language and linguistic universals, Sandler and Lillo-Martin (2006) discuss the phonetic/phonological universals shared by signed and spoken languages. These include "a sublexical level of structure that is meaningless, discrete, finite, and systematically organized" (272). For all signed languages, this phonological level makes use of hand configuration and location, originally characterized by Stokoe (1960), as well as movement. Sandler and Lillo-Martin (2006) put forth prosodic structure as another modality-independent universal, as well as patterns involving articulatory features that cluster by articulator, reflecting the modality. A difference is that phonological structure tends to be simultaneous in signed languages but sequential in spoken language, and Sandler and Lillo-Martin (2006) argue that the proliferation of phonological rules in spoken language is due to the predominance of (rapid) sequential structure in spoken languages. For more discussion of mechanistic and ecological differences related to signed vs. spoken modality, see Meier et al. (2002).

The Konstanz Universals Archive (Plank, 2000) is a database of linguistic universals proposed in the literature (including many that have been falsified). Hyman (2008) commented on 487 of the proposed universals that he found to refer to phonology or phonetics. Other resources used to explore phonological universals of spoken language are the Lyon-Albuquerque phonological systems database (LAPSyd, Maddieson 2014), an inventory database that serves as the successor to UPSID, the inventory database PHOIBLE (Moran et al., 2014), and P-base (Mielke and Brohan, 2018), which is a database of phonological alternations and distributional restrictions. Databases for signed languages are SignPhon (Crasborn et al., 2001) and the Sign Language Analyses Database (SLAY, Tatman 2015).

There have been many conference sessions and books broadly focused on phonetic and/or phonological universals among spoken languages. These include Greenberg (1963), Greenberg (1966a), Greenberg (1978), which includes fourteen chapters by various authors all on various topics within phonetic and phonological universals, Mairal and Gil (2006), Butterworth et al. (1983), and the nine papers from a special session at the 9th International Congress of Phonetic Sciences (ICPhS) (see Ohala 1979).

Blevins, Juliette. 2004. Evolutionary Phonology. Cambridge: Cambridge University Press.

Booij, Geert E. 1983. Principles and parameters in prosodic phonology.

Butterworth, Brian, Bernard Comrie, and Östen Dahl, ed. 1983. Explanations for language universals. The Hague: Mouton.

Bybee, Joan, and Shelece Easterday. 2022. Primal consonants and the evolution of consonant inventories. Language Dynamics and Change 1:1–33.

Cahill, Mike. 2007. More universals of tone. Ms. SIL. Downloadable at: http://orthographyclearinghouse.org/papers/mikeCahillonUniversalsofTone.pdf

Cho, Taehong, and Peter Ladefoged. 1999. Variation and universals in VOT: evidence from 18 languages. Journal of Phonetics 27:207–229.

Chodroff, Eleanor, Alessandra Golden, and Colin Wilson. 2019. Covariation of stop voice onset time across languages: Evidence for a universal constraint on phonetic realization. The Journal of the Acoustical Society of America 145:EL109–EL115.

Chomsky, Noam, and Morris Halle. 1968. The Sound Pattern of English. New York: Harper and Row.

Crasborn, Onno A, Harry van der Hulst, and Els van der Kooij. 2001. Signphon: a phonological database for sign languages. Sign language & linguistics 4:215–228.

Duanmu, San. 2016. A theory of phonology features. Oxford: Oxford University Press.

Evans, Nicholas, and Stephen C. Levinson. 2009. The myth of language universals: Language diversity and its importance for cognitive science. Behavioral and Brain Sciences 32:429–448.

Flemming, Edward. 2004. Contrast and perceptual distinctiveness. In Phonetically-based phonology , ed. B. Hayes, R. Kirchner, and D. Steriade, 232–276. Cambridge University Press.

Flemming, Edward S. 2002. Auditory representations in phonology. New York: Routledge.

Gick, Bryan, Fiona Campbell, Sunyoung Oh, and Linda Tamburri-Watt. 2006. Toward universals in the gestural organization of syllables: A cross-linguistic study of liquids. Journal of Phonetics 34:49–72.

Greenberg, Joseph H., ed. 1963. Universals of language. Cambridge, Massachusetts: MIT Press.

Greenberg, Joseph H. 1966a. Language universals: With special reference to feature hierarchies. The Hague, Paris: Mouton & Co.

Greenberg, Joseph H. 1966b. Synchronic and diachronic universals in phonology. Language 42:508–517.

Greenberg, Joseph H., ed. 1978. Universals of human language, volume 2: Phonology. Stanford, California: Stanford University Press.

Hall, Daniel Currie. 2011. Phonological contrast and its phonetic enhancement: Dispersedness without dispersion. Phonology 28:1–54.

Hyman, Larry M. 2008. Universals in phonology. The Linguistic Review 25:83–137.

Hyman, Larry M., and Russel G. Schuh. 1974. Universals of tone rules: evidence from West Africa. Linguistic Inquiry 5:81–115.

Jaeger, Jeri J. 1978. Speech aerodynamics and phonological universals. In Annual Meeting of the Berkeley Linguistics Society, volume 4, 312–329.

Javkin, Hector Raul. 1978. Phonetic universals and phonological change. Doctoral Dissertation, University of California, Berkeley.

Kawasaki, Haruko. 1982. An acoustical basis for universal phonotactic constraints. Doctoral Dissertation, University of California, Berkeley.

Kawasaki-Fukumori, Haruko. 1992. An acoustical basis for universal phonotactic constraints. Language and Speech 35:73–86.

Keating, Patricia. 1984a. Physiological effects on stop consonant voicing. UCLA Working Papers in Phonetics 59:29–34.

Keating, Patricia. 1984b. Universal phonetics and the organization of grammars. UCLA Working Papers in Phonetics 59:35–49.

Kiparsky, Paul. 2018. Formal and empirical issues in phonological typology. In Phonological typology, ed. Larry Hyman and Frans Plank, 54–106. Berlin: Mouton de Gruyter Phonetics and Phonology series.

Ladefoged, Peter. 2002. Speculations on the deeper causes of phonetic universals. UCLA Working Papers in Phonetics 35–49.

Liljencrants, Johan, and Bjorn Lindblom. 1972. Numerical simulation of vowel quality systems: the role of perceptual contrast. Language 48:839–862.

Lindblom, Björn, Peter MacNeilage, and Michael Studdert-Kennedy. 1983. Self-organizing processes and the explanation of phonological universals. In Explanations for language universals, ed. Brian Butterworth, Bernard Comrie, and Östen Dahl, 181–203. The Hague: Mouton.

Lindblom, Björn, and Ian Maddieson. 1988. Phonetic universals in consonant systems. In Language, Speech, and Mind: Studies in Honour of Victoria Fromkin, ed. Larry M. Hyman and Charles N. Li, 62–78. London: Routledge.

Maddieson, Ian. 1978. Universals of tone. In Universals of human language, volume 2: Phonology, ed. Joseph H. Greenberg, 335–366. Stanford, CA: Stanford University Press.

Maddieson, Ian. 1997. Phonetic universals. In The handbook of phonetic sciences, ed. William J. Hardcastle and John Laver, 619–639. Oxford and Cambridge: Blackwell.

Maddieson, Ian. 2014. LAPSyD: Lyon-Albuquerque phonological systems database. CNRS, Lyon, France. URL https://lapsyd.huma-num.fr/lapsyd/.

Maddieson, Ian, and Kristin Precoda. 1990. Updating UPSID. In UCLA Working Papers in Phonetics, volume 74, 104–111.

Mairal, Ricardo, and Juana Gil. 2006. Linguistic universals. Cambridge University Press.

Meier, Richard P., Kearsy Cormier, and David Quinto-Pozos. 2002. Modality and structure in signed and spoken languages. Cambridge University Press.

Mielke, Jeff, and Anthony Brohan. 2018. P-base 3.0. URL https://pbase.phon.chass.ncsu.edu/, database of sound patterns.

Moran, Steven, Daniel McCloy, and Richard Wright. 2014. Phoible online. URL http://phoible.org, Leipzig: Max Planck Institute for Evolutionary Anthropology, Accessed on 2022-02-16.

Moreton, Elliott. 2008. Analytic bias and phonological typology. Phonology 25:83–127.

Ohala, John J. 1979. Phonetic universals in phonological systems and their explanation: summary of moderator’s introduction. In Proceedings of the 9th International Congress of Phonetic Sciences, volume 2, 5–8. The papers are available at https://www.coli.uni-saarland.de/groups/BM/phonetics/icphs/ICPhS1979/ICPhS1979.html

Ohala, John J. 2011. Accommodation to the aerodynamic voicing constraint and its phonological relevance. In Proceedings of ICPhS XVII , 64–67.

Plank, Frans. 2000. Universals archive. URL https://typo.uni-konstanz.de/rara/category/universals-archive/

Plauché, Madelaine C, Rosemary Beam De Aconza, Rungpat Roengpitya, and William F Weigel. 1998. Glottalized sonorants: A phonetic universal? In Annual Meeting of the Berkeley Linguistics Society, volume 24, 381–390.

Sandler, Wendy. 2017. The challenge of sign language phonology. Annual review of Linguistics 3:43–63.

Sandler, Wendy, and Diane Lillo-Martin. 2006. Sign language and linguistic universals. Cambridge: Cambridge University Press.

Solé, Maria-Josep. 2007. Controlled and mechanical properties in speech. Experimental approaches to phonology 302–321.

Stokoe, William. 1960. Sign language structure: An outline of the visual communication of the American deaf. In Studies in linguistics, occasional papers 8.

Tatman, Rachael. 2015. The sign language analyses (slay) database. Technical report, University of Washington Working Papers in Linguistics.

Vogel, Irene. 2009. Universals of prosodic structure. Universals of language today 59–82.

Westbury, John, and Patricia Keating. 1986. On the naturalness of stop consonant voicing. Journal of Linguistics 22:145–166.