Content uploaded by Camilla Power
Author content
All content in this area was uploaded by Camilla Power on Aug 20, 2019
Content may be subject to copyright.
Metadata of the chapter that will be visualized online
Chapter Title Ritual and Speech Coevolution
Copyright Year 2018
Copyright Holder Springer International Publishing AG, part of Springer Nature
Corresponding Author Family Name Power
Particle
Given Name Camilla
Suffix
Division/Department Department of Anthropology
Organization/University University of East London
City London
Country UK AU1
Email c.c.power@uel.ac.uk
1R
2Ritual and Speech
3Coevolution
4Camilla Power
5Department of Anthropology
6University of East London, London, UKAU1
7Synonyms
8Conventionalization;Metaphor;Mimesis;Sym-
9bolic culture
10 Definition
11 Thesis originally developed by anthropologist
12 Chris Knight on the important consideration of
13 novel forms of group cooperation, and
14 corresponding conflict, in the emergence of
15 human language, analogous to signaling in other
16 species.
17 Introduction
18 In this entry, I discuss the ritual/speech coevolu-
19 tion hypothesis for the evolutionary emergence of
20 spoken language. This brings the main theory on
21 the evolution of animal signals to bear on the
22 question of language origins. Our closest primate
23 relatives are fundamentally constrained in their
24 ability to mimic or produce novel vocalizations,
25under pressure to maintain reliability in their sig-
26nals. The ritual/speech coevolution theory
27explains how those selection pressures were
28relaxed in the case of human evolution through
29the generation of novel levels of in-group trust.
30Main Text AU2
31Ritual/speech coevolution is a hypothesis
32explaining the evolutionary emergence of spoken
33language. It rests on the assumption that language
34can only evolve given novel conditions of social
35cooperation widespread across human communi-
36ties –effectively cooperation between strangers.
37In the contexts of human evolution, the only
38medium for securing such cooperation within
39and between groups was costly ritual
40(cf Durkheim 1912; Knight et al. 1995; Maynard
41Smith and Szathmary 1995; Deacon 1997;
42Rappaport 1999; Irons 2001; Alcorta and Sosis
432005; Knight and Lewis 2017).
44Ritual/speech coevolution brings to the prob-
45lem of language origins the main body of theory
46applied to the evolution of animal signals. Signal
47evolution theory deals with the emergence, func-
48tion, and design of animal signals (Maynard
49Smith and Harper 2003). Focused on social inter-
50action and behavior in the world, signal evolution
51is necessarily political in its approach. Central
52debates concern the honesty and reliability of
53signals (Zahavi and Zahavi 1997), manipulation
54and mindreading (Krebs and Dawkins 1984), and
#Springer International Publishing AG, part of Springer Nature 2018
T. K. Shackelford, V.A. Weekes-Shackelford (eds.), Encyclopedia of Evolutionary Psychological Science,
https://doi.org/10.1007/978-3-319-16999-6_3317-1
55 the effects of shared versus conflicting interests in
56 outcomes (Maynard Smith and Harper 2003).
57 Among primates generally, tactical use of
58 deception is predicted by neocortex volume
59 (Byrne and Corp 2004), a finding consistent with
60 the Machiavellian intelligence hypothesis –the
61 idea that intelligence evolves under pressure to
62 deal with social challenges. Humans are clearly
63 Machiavellian in terms of political alliance forma-
64 tion and capacities for deception. Issues of trust
65 and reliability are critical to any understanding of
66 language origins.
67 While apes and monkeys have rich repertoires
68 of calls, they seem to be severely restricted in their
69 ability to mimick or to produce novel vocaliza-
70 tions (Fitch and Zuberbühler 2013). In particular,
71 when phonating, they lack neural control over
72 articulators such as the lips, tongue, soft palate,
73 etc. This means that our closest relatives are
74 unable to manipulate vocalizations at will. No
75 monkey or ape has an inflexible tongue, but
76 when communicating it leaves the tongue out.
77 Knight and Lewis (2017) draw on signal evolu-
78 tion theory to explain why. While sound –unlike
79 visible gesture –carries over distances, goes
80 around corners, and works in the dark, in these
81 contexts listeners lack corroborating evidence of
82 reliability. It makes sense then that vocalizations
83 tied to bodily and emotional states will be per-
84 ceived as more reliable than those which can be
85 altered at will. Mistrusting one another’s schem-
86 ing, Machiavellian minds, primates ignore the all
87 too-flexible tongue, preferring to rely on the evi-
88 dence of their own eyes and ears. In this view,
89 apes are “too clever for words”(Knight 1998: 72).
90 So the conundrum of speech and language
91 evolution is to explain how and why natural selec-
92 tion, in the human case, switched from quarantin-
93 ing the primate tongue –excluding it from all but a
94 marginal communicative role –to developing and
95 fine-tuning that same tongue’s role as the most
96 important speech articulator of all.
97 Signal evolution theory contrasts two diver-
98 gent trajectories depending on the degree to
99 which signaler and receiver share interest in the
100 same outcomes. Where we find high-cost, repeti-
101 tive, multimedia displays, we may infer a function
102 in terms of social manipulation, conflict, and
103exploitation. Resistance by receivers sets up
104selection pressures acting on signal design. Sig-
105nalers who encounter “sales resistance”are driven
106to respond by prolonging and repeating signals,
107increasing amplitude, and resorting to costly mul-
108timedia displays. Such extravagant advertisers
109include peacocks displaying to would-be mates
110or caribou bulls bellowing in the rut. Zahavi
111(1987) shows how the discernible costs of such
112displays enhance their credibility by tapping and
113hence testing the very reservoirs of quality that
114signalers are attempting to advertise. The evolu-
115tion of high-cost signaling is driven by skeptical
116receivers pushing signalers to ever greater com-
117petitive effort to prove their quality.
118By contrast, where we find low-cost, quiet, and
119efficient signals, a cooperative audience can be
120inferred. If signalers can afford to cut their emis-
121sion costs, it is only because listeners are investing
122corresponding effort in receiving, decoding, and
123acting on signals. This implies that signalers and
124receivers significantly share interests. For such
125“conspiratorial whispering”(Krebs and Dawkins
1261984) to evolve, signalers must be imparting use-
127ful information to receivers and those receivers
128are not expecting to be deceived.
129The ultimate cost-cutting strategy would be to
130resort to purely tokenistic, wholly conventional
131signals that can be processed categorically at
132speed –relieving listeners of the need to evaluate
133gradations in physical performance and signalers
134of such costly performance. According to Zahavi
135(1993), however, animal “conspiracies”are never
136sufficiently cooperative. Internal conflict and
137skepticism precludes selection for reliance on
138tokenistic “paper money.”Nowhere in the living
139world do we find purely conventional signaling –
140with the puzzling exception of human speech.
141Turning to human symbolic communication,
142Knight (1998,1999) argued that our ancestors
143developed divergent, formally contrastive types
144of communication along these two “high-cost”
145and “low-cost”trajectories: ritual and speech.
146Table 1shows the diametric opposition of
147characteristics of each. On the basis of signal
148evolution theory, we can infer that speech
149emerged in a cooperative context while ritual did
150not. For speech to have evolved, “conspiratorial
2 Ritual and Speech Coevolution
151 whispering”in the human case must have been
152 anomalously trusting. By contrast, ritual –with its
153 costly, inefficient features of redundancy and
154 display –can only have emerged from a dynamic
155 of conflict, manipulation, and exploitation.
156 The paradox here, in the light of nonhuman
157 primate vocalization, is that the extremely low-
158 cost, conventional codes of speech can easily
159 leave listeners vulnerable to deception. With no
160 intrinsic link between sounds and their purely
161 arbitrary meanings, words are routinely
162 decoupled from emotional veracity or real-world
163 stimuli. Despite speech having no intrinsic reli-
164 ability, human conversation works on Grice’s
165 “cooperative principle”(Grice 1969), with partic-
166 ipants, even where they may be in some degree of
167 social conflict, cooperating at the level of mutual
168 conversational ends. If humans are on “speaking
169 terms,”they expect intentional honesty. Somehow
170 this new default of honesty –honesty in the
171 deployment of volitional, conventional signals –
172 became established. In the course of human evo-
173 lution what were once frequency-dependent tacti-
174 cal deceptions became increasingly harnessed to a
175 reversed social function: group-wide sharing of
176 socially useful information.
177 This paradox can be unraveled if we consider
178 that whereas speech is fundamentally interper-
179 sonal, ritual operates group-on-group and may
180 function to demarcate boundaries of those groups
181 (Cohen 1985; Harrison 1993). Experience of the
182 high-cost, multimedia signals involved in ritual
183will differ according to whether the receiver is
184outside or inside the performing group. For out-
185siders, the costly signals are manipulative and
186exploitative, needing to overcome “sales-
187resistance”by impressing observers with the qual-
188ity of performance. Among insiders, the “collec-
189tive effervescence”(Durkheim 1912), aroused by
190singing, dancing, and musicking together, will
191intensify in-group trust and generate a sense of
192group identity as “We”(Rappaport 1999). Focus-
193ing on the emotional responses of individuals to
194ritual experience, Alcorta and Sosis (2005) show
195that the very costliness of undergoing ritual
196enables those individuals to demonstrate in hard-
197to-fake terms their commitment to the group.
198There are two major emergent effects from
199such proto-ritual experience. First, the novel
200levels of in-group trust can now permit conven-
201tionalization and shorthand reference –volitional,
202conventional signals –in communicating among
203group members, now on the way to becoming a
204speech community. Secondly, in representing
205their coherent and enduring solidarity, ritual per-
206formers generate sacred, supernatural, and
207counter-intuitive concepts –the first gods
208(Durkheim 1912; Knight et al. 1995; Rappaport
2091999; Alcorta and Sosis 2005).
210Roy Rappaport analyzes the form and features
211of ritual that make it “the social act basic to
212humanity”(Rappaport 1999: 31). Ritual over-
213comes the divisiveness of individually held
214beliefs. Defining ritual as “the performance of
t1:1Ritual and Speech Coevolution, Table 1 Speech versus ritual (After Knight 1999: 231)
Speech Ritual
t1:2
Cheap signals Costly signals
t1:3
Interpersonal Group-on-group
t1:4
Two-way communication One-way signals
t1:5
Low amplitude High amplitude
t1:6
Dispassionate Emotive
t1:7
Vocal-auditory Multi-media
t1:8
Digital Analog
t1:9
Discrete-combinatorial Holistic
t1:10
Productivity/creativity Repetition/redundancy
t1:11
Stress on novelty Stress on conservatism
t1:12
Conventionally coded Iconic and indexical
t1:13
Focus on underlying intentions Focus on body-boundaries and surfaces
t1:14
Ritual and Speech Coevolution 3
215 more or less invariant sequences of formal acts
216 and utterances not entirely encoded by the per-
217 formers”(Rappaport 1999: 24), Rappaport
218 focuses on the unique conjunction of features in
219 ritual. Ritual form permits the intersection of the
220 symbolic with what is not purely symbolic. Ritual
221 combines formal canonical messages –encoded
222 by other than the performers –with the indexical
223 signals, the current state, mood, and emotions of
224 the flesh and blood performers. This “substantia-
225 tion of form and ...informing of substance”mean
226 that the establishment of convention is intrinsic to
227 ritual. Without this, no social contract, moral
228 order, or language could exist.
229 The function of ritual is not to differentiate
230 between lexical meanings but to establish, for
231 everyone, an overarching meaning –a meta-
232 performative or Word –from whose subsequent
233 fragmentation a limitless multiplicity of subsidi-
234 ary meanings can be derived. Rappaport explains
235 how apparently irrational nonsense –perhaps the
236 endless repetition of just a few meaningless
237 sounds –may “provide the ground, deeper than
238 logic and beyond logic’s reach”upon which to
239 establish sufficient collective authority and
240 mutual trust to build up “the usages and rules of
241 social life,”in turn enabling words to make sense.
242 While Rappaport did not have any very satis-
243 factory model for the evolution of ritual (hence
244 speech), Knight and Lewis (2017), within the
245 ritual/speech coevolution paradigm, provide a
246 specific evolutionary account. Words and gram-
247 mar, they claim (Knight and Lewis 2017: 435), are
248 “means of navigating within a shared virtual
249 world. Singing, dancing, and other forms of com-
250 munal ritual are necessary to join people together
251 in such ideal or imagined worlds.”Language then
252 will not even begin to evolve unless ritual action
253 has already begun to establish intensified levels of
254 community-wide trust in association with a shared
255 virtual domain. This explains why nonhuman pri-
256 mates, confined as they are to the brute world of
257 physical facts, have no need of it.
258 Knight and Lewis explain how the generative
259 principle for both words and grammar is the
260 ostensive-inferential process underlying meta-
261 phor. In this view, symbolic communication
262 “rests on the ability of listeners to infer relevant
263communicative intentions from expressions that,
264interpreted literally, are inadequate or untrue”
265(Knight and Lewis 2017: 436). Not only does
266language comprise zero or very low-cost voli-
267tional vocal signals, but at the level of both
268words and rules, lexicon, and grammar, these
269amount to conventionally agreed “lies”or decep-
270tions. Whereas nonhuman primates would simply
271reject these sound sequences, we humans instead
272search out the hidden implications and communi-
273cative intentions.
274Knight and Lewis draw on hunter-gatherer eth-
275nography to work out a model for the emergence
276of such socially agreed “deceptions.”Human abil-
277ities to manipulate pitch were first exercised in
278fooling outsiders: animals, whether these were
279prey being hunted (by men) or predators being
280evaded (by women and children). Song first –a
281communal polyphony of meaningless sounds –
282used to confuse and deter predators regarding
283numbers of hominins in groups, would have
284developed capacities for intentionally varying
285pitch. In such a context, the emotional solidarity
286among chorusing groups would override any dan-
287gers of deceptive use. Shorthand snatches of song,
288animal-sound mimickry, or pantomime dance
289steps could now begin to take on specific shared
290meaning. From there, a fully grammaticalized
291language could emerge with extraordinary speed,
292as long as the freedom to innovate –freedom to
293“say”one thing while “meaning”another –was
294maintained. What had blocked earlier hominins
295and hominids from any such grammaticalization
296process was “the burden imposed on all signals to
297incorporate some costly component to demon-
298strate reliability”(Knight and Lewis 2017: 445).
299Grammaticalization is based on a tendency to
300efficiency, but requirement of reliability in signals
301demands the opposite, increased cost. The release
302from this constraint frees the tongue and other
303articulators so they can be recruited to increas-
304ingly efficient, conventionalized communication:
305speech.
306Knight and Lewis (2017) aim to make their
307model testable by specifying, in detail, the
308“world’sfirst metaphor,”identifying it as a gen-
309dered ritual performance in which the core princi-
310ples of primate politics are overturned. They argue
4 Ritual and Speech Coevolution
311 that as female hominin ancestors faced increasing
312 reproductive costs with very large-brained off-
313 spring, critical points of conflict occurred when-
314 ever an alpha male attempted to monopolize an
315 imminently fertile (menstruating) female. This
316 would have conflicted sharply with reproductive
317 interests of both other females and other males.
318 Ritual display developed as female coalitions
319 mounted resistance to any would-be alpha, with
320 the whole female group signaling their non-
321 availability to males through song and pantomime
322 dance: we are the wrong species (animals); we are
323 the wrong sex (males); and this is the wrong time
324 (blood). By identifying their own “menstrual”
325 blood as the blood of the game animals they
326 wanted the men to hunt, women’sfirst morally
327 authoritative ritual construct established categoric
328 rules encompassing sex, kinship, and economics.
329 The specific model proposed here has the merit
330 of testability. But the general ritual/speech coevo-
331 lution paradigm has advantages (Power 2014).
332 Firstly it offers a coherent basis for continuity
333 with animal communication systems through the-
334 ory of ritualized behavior. Secondly, regarding the
335 archaeological record, it makes the clear predic-
336 tion that the origins of language will be marked by
337 evidence for ritual behavior (Watts 2014).
338 Cross-References
339 ▶Communication, Cues, and Signals
340 ▶Cooperative Coalitions
341 ▶Evolution of Cooperation
342 ▶Female Resistance
343 ▶Grammaticalization
344 ▶Handicap
345 ▶Manipulation and Dishonest Signals
346 ▶Reliability and Deception in Language
347 ▶Sexual Conflict Theory
348 ▶Signal Reliability.
349References
350Alcorta, C. S., & Sosis, R. (2005). Ritual, emotion and
351sacred symbols. Human Nature, 16, 323–359.
352Byrne, R. W., & Corp, N. (2004). Neocortex size predicts
353deception rate in primates. Proceedings of the Royal
354Society of London B, 271, 1693–1699.
355Cohen, A. P. (1985). The symbolic construction of commu-
356nity. London: Tavistock.
357Deacon, T. (1997). The symbolic species: The co-evolution
358of language and the human brain. London: Penguin.
359Durkheim, E. (1912). Les formes élémentaires de la vie
360religieuse. Paris: Alcan.
361Fitch, W. T., & Zuberbühler, K. (2013). Primate precursors
362to human language: Beyond discontinuity. In
363E. Altenmüller, S. Schmidt, & E. Zimmermann (Eds.),
364Evolution of emotional communication (pp. 26–48).
365Oxford: Oxford University Press.
366Grice, H. P. (1969). Utterer’s meanings and intentions.
367Philosophical Review, 78, 147–177.
368Harrison, S. (1993). The mask of war. In Violence, ritual
369and the self in melanesia. Manchester & New York:
370Manchester University Press.
371Irons, W. G. (2001). Religion as hard-to-fake sign of com-
372mitment. In R. Nesse (Ed.), Evolution and the capacity
373for commitment (pp. 292–309). New York: Russell
374Sage Foundation.
375Knight, C. (1998). Ritual/speech co-evolution: A solution
376to the problem of deception. In J. R. Hurford,
377M. Studdert-Kennedy, & C. Knight (Eds.), Approaches
378to the evolution of language. Social and cognitive bases
379(pp. 68–91). Cambridge: Cambridge University Press.
380Knight, C. (1999). Sex and language as pretend-play. In
381R. Dunbar, C. Knight, & C. Power (Eds.), The evolu-
382tion of culture (pp. 228–247). Edinburgh: Edinburgh
383University Press.
384Knight, C., & Lewis, J. (2017). Wild voices: Mimicry,
385reversal, metaphor, and the emergence of language.
386Current Anthropology, 58, 435–453.
387Knight, C., Power, C., & Watts, I. (1995). The human
388symbolic revolution: A darwinian account. Cambridge
389Archaeological Journal, 5,75–114.
390Krebs, J. R., & Dawkins, R. (1984). Animal signals: Mind-
391reading and manipulation. In J. R. Krebs &
392N. B. Davies (Eds.), Behavioural ecology: An evolu-
393tionary approach (2nd ed., pp. 380–402). Oxford:
394Blackwell Scientific Publications.
395Maynard Smith, J., & Harper, D. (2003). Animal signals.
396Oxford: Oxford University Press.
397Maynard Smith, J., & Szathmary, E. (1995). The major
398transitions in evolution. Oxford: W. H. Freeman.
399Power, C. (2014). The evolution of ritual as a process of
400sexual selection. In D. Dor, C. Knight, & J. Lewis
401(Eds.), The social origins of language (pp. 196–207).
402Oxford: Oxford University Press.
Ritual and Speech Coevolution 5
403 Rappaport, R. A. (1999). Ritual and religion in the making
404 of humanity. Cambridge: Cambridge University Press.
405 Watts, I. (2014). The red thread: Pigment use and the
406 evolution of collective ritual. In D. Dor, C. Knight, &
407 J. Lewis (Eds.), The social origins of language
408 (pp. 208–227). Oxford: Oxford University Press.
409 Zahavi, A. (1987). The theory of signal selection and some
410 of its implications. In V. P. Delfino (Ed.), International
411symposium of biological evolution (pp. 305–327). Bari:
412Adriatic Editrice.
413Zahavi, A. (1993). The fallacy of conventional signalling.
414Philosophical Transactions of the Royal Society of
415London, 340, 227–230.
416Zahavi, A., & Zahavi, A. (1997). The handicap principle.
417New York and Oxford: Oxford University Press.
6 Ritual and Speech Coevolution
Author Queries
Encyclopedia of Evolutionary Psychological Science
Chapter No.: 3317-1
___________________________________________________________________
Query Refs. Details Required Author's response
AU1 Please be aware that your name and affiliation and if
applicable those of you co-author(s) will be published as
presented in this proof. If you want to make any
changes, please correct the details now. Note that
corrections after publication will no longer be possible.
AU2 Please provide appropriate heading instead of "Main
Text."
Note:
If you are using material from other works please make sure that you have obtained the necessary permission from
the copyright holders and that references to the original publications are included.
