Rebus and acrophony in invented writing

ABSTRACT

Rebus and acrophony are crucial in the development of ancient invented scripts from Mesopotamia (cuneiform), China, Mesoamerica (Maya), Egypt, and scripts which may have been created through exposure to literate cultures (Anatolian Hieroglyphic and Nahuatl). Yet, these two linguistic mechanisms have been understudied from a terminological, contextual and comparative perspective. This article aims to address issues regarding their definition, development and application in script formation. The scope of our study is all attested writing systems that are largely iconic in their sign repertoire, and whose phonetic values were generated anew based on an underlying language (hence ‘invented’). This allows us to chart how writing systems are created ex novo and what trajectories of development are put into practice when phonetisation takes place. We show some reliable patterns of universal mechanisms, observable from a comparative perspective. We also demonstrate that these patterns attest to a verifiable degree of phonological awareness that ties the process of phonetisation to the path to script formation. We further highlight that the tendencies discerned from deciphered writing systems provide ways to test hypotheses in the study of iconic writing systems which are undeciphered, such as the Indus Valley script and the Rongorongo of Easter Island.

KEYWORDS:

• Rebus
• acrophony
• phonetisation
• script formation

Acknowledgements

Probably in the same way that ancient scripts took shape, this article is the result of cumulative effort, the fruit of constant discussion, and the product of regular tweaking and re-thinking. We are grateful to Ignasi-Xavier Adiego, Alex de Voogt, Christian Prager, and Brent Davis, for their many references, ideas, and support. We also thank Lorenzo Lastilla for drawing the map. In the initial stages of analysis, we relied on the enthusiasm and knowledge of Erik Boot, who guided us through the meanders of the Maya signs. This article is dedicated to his memory.

Disclosure statement

No potential conflict of interest was reported by the author(s).

ORCID

Miguel Valério http://orcid.org/0000-0002-9952-2057

Notes

1 This excludes cases of more recent scripts sometimes defined as “invented”, but whose sign shapes or values were certainly or potentially inspired by other writing systems: among others, Pahawh Hmong (Laos), Vai (Liberia) and Cherokee (United States).

2 Seyffarth used the Latin adjective acrophonicum and the German equivalent akrophonisch. The term appeared around the same time as acrologique, a coinage by Klaproth (1827, p. 33–34, n. 2), another detractor of Champollion. Klaproth used ‘acrologic’ to describe the views of Goulianov, a scholar who believed the hieroglyphs were logographs that evoked the initial sound of the word they meant to notate (as if e.g. one drew a sign depicting a pig to write ‘plant’ or ‘pain’, etc.). ‘Acrology’ and ‘acrologic’ were used as synonymous with ‘acrophony’ and ‘acrophonic’ throughout the 19th century until the latter eventually became dominant.

3 First used in this sense by Steinthal (1852, p. 113). Here we provide the reconstructed Old Germanic version of the letter names, after Williams (2004, p. 263).

4 We shall not address whether proto-cuneiform was used by speakers of a non-Sumerian language (Englund, 1998). There are clear matches between early cuneiform sign shapes, sign values and Sumerian words which imply phonetisation based on that language. The residual involvement of another remains a possibility, but so far there is no consensual evidence.

5 The two earliest examples of writing are Monument 13 from La Venta (in the Olmec heartland) and the Monument 3 from San José Mogote (Oaxaca, Zapotec Highlands). These would be instances of early Olmec and Zapotec ‘scripts’, respectively. However, these inscriptions are not deciphered and we cannot tell whether phonetic notation was involved. A possible third case from La Venta, Tabasco (Pohl, Pope, & von Nagy, 2002), dating to ca. 650 BCE and associated as well with the Olmecs by archaeologists, seems inconclusive.

6 The relationship between the two scripts remains unclear, as do their alleged ties to the so-called Zapotec ‘script’ (Macri & Looper, 2003, pp. 4–5). It has been suggested that Maya derives from Isthmian (Lacadena, 2010), but it is also possible that it was the other way around (Boot, 2010, p. 157, fn. 34). Curiously, Isthmian sign shapes (as attested) seem more schematised, and in general the script is less iconography-dependent than Maya (Houston, 2004, pp. 297–298).

7 Or similar. ‘Basket’ is the translation by Bóttero (2004, p. 251), based on the shape of the sign in the oracle bone inscriptions form, whereas Baxter (1992, p. 28, 472) has ‘winnowing basket’, and Baxter and Sagart (2014, p. 343) have ‘base’. The latter corresponds better to shape of the sign on the bronze inscriptions, (a container on a stand), which is the predecessor of modern jī.

8 Another strategy existed by which a sign became polyvalent. It consisted of representing the word depicted by the sign, and, metonymically, a closely associated concept: e.g. the graph ‘mouth’ could also denote the verb ‘to call, name’ (Boltz, 1986, p. 426). Evidently, this procedure is semantic and hence is of no consequence for our purposes.

9 For example, r was used to write syllables with different vocalisms in words such as nṯrt /naṯārat/ ‘goddess’, nṯrj /nuṯrij/ ‘divine (m.)’, and nfr /nāfir/ ‘good, beautiful’ (Loprieno, 1995, pp. 35, 39, 53). In this way, r denoted strictly the consonant /r/.

10 This is true even for triconsonantals derived from logographs denoting more than one root. For example, as a logogram and determinative, expresses the concept of ‘thigh (i͗w^c ) or leg (swt) of beef’ and, by metaphor, also the idea of ‘reward’ (i͗sw). Thanks to this ambivalence, the sign developed two separate phonetic values, i͗w^c and i͗sw. As a result, it spelled words such as i͗w^c ‘to inherit’ and i͗w^c t ‘inheritance’, but also i͗sw ‘reward’, i͗swt ‘representative’ and i͗swj ‘testicles’ (Edgerton, 1940, p. 484; Hoch, 1998, p. 22).

11 Perhaps their vocalism diverged, as it is accepted that both uniconsonantals and biconsonantals omitted vowels. Conversely, in cases like ḥr /ħar/ ‘FACE’ → ḥr /ħar/ ‘on’ (Loprieno, 1995, pp. 56, 100; Hoch, 1998, p. 9) the reconstructed vocalisations suggest that the logogram ‘FACE’ was used phonetically thanks to close homophony. In conclusion, since the vocalism of most ancient Egyptian words is irrecoverable, we cannot be sure of the extent to which rebus could ignore the vowels of biconsonantals.

12 This principle was applied even in later acrophonic reforms of the script, for example when the ‘traditional’ biconsonantal rw came to be used as r in the New Kingdom (Vernus, 2016, p. 157–158). In fact, j, w and ȝ were occasionally omitted from spellings altogether.

13 Of course, language may not have been the only determining factor. In many cases (but not all and not systematically), the simplicity of the iconic logograms in terms of design may have been key: hieroglyphs with less strokes and complex details like r, ḳ, t, k, s , p, b, and h are among the easiest to inscribe. Since they would become the phonetic complements to use and repeat in the spelling of several words, economy of effort may have been considered.

14 Ten of the later 26 standard uniconsonantals had already been created by the time of Narmer: i͗, p, f, n, r, ḥ, s, t, ṯ, and ḏ (Kahl, 2004, p. 119). Moreover, the uniconsonantal n is attested already in the reign of Sekhen/Ka (ca. 3150 BCE), as a complement to the logogram (‘lotus’ nḥbt) in the spelling of nḥb ‘gift’.

15 We find redundant orthographies of the type nfr+f-r-t + semantic complements for nfrt ‘goodness, good things’ (Hoch, 1998). Parts of the root, not just the affix -t, were phonetically expressed (but cf. nfr+t + semantic complements for nfrt ‘cattle’).

16 According to Justeson (2012), a few CVC syllabograms occur in Maya: e.g. nah, spelling both naah ‘house’ and nah ‘first’. However, it is important (though not always easy) to distinguish between rebus as a mechanism for deriving phonetic values at the formative stages of the script, and rebus as an orthographic device throughout the life of the script.

17 Sign T671 chi, depicting a hand with the thumb touching the forefinger, stands logographically for day 7 or manik, which was a day named ‘deer’ throughout Mesoamerica. While ‘deer’ is *chij in Proto-Ch’olan (< Proto-Mayan *keeh), the shape of the sign cannot have motivated its phonetic value. Hence Law et al. (2014) argue that the syllabic value chi was based on the calendrical value of the sign but emerged secondarily only after the palataliation of *keeh ‘deer’ had occurred. There is a variant sign T- = AV1, which depicts a deer head, but it is reported by Macri and Looper (2003, p. 84) to be a substitute for chi in throne titles (presumably a later addition to the script).

18 The phonetic component of the system has been deemed underdeveloped, as glyphs for the frequent syllables ti and qui are unattested (Whittaker, 2009), and in general it still relied heavily on puns. Yet certain glyphs may be missing due to the limitations of the corpus, while Maya and other logo-phonetic scripts made also extensive use of puns. Likewise, partial spellings in Nahuatl (e.g. MATLA for Matlatzinco; Lacadena, 2008, p. 5, fig. 1-e) are comparable to spellings in the formative stages of other scripts, for instance Anatolian Hieroglyphic.

19 Two linguistic details are worth mentioning. First, Nahuatl tl represents a lateral affricate /tɬ/, not a consonant cluster. Second, the suffix -tl (and its allomorphs) is not “absolutive” in the sense of marking the absolute case as part of an ergative system, but rather in the sense of marking the dictionary form, as described.

20 At times, the interpretations of the origin of signs can point either to invention or adaptation. Paradigmatic is the pair comprising T- = YSA mu, depicting a frog(?) with a curl (and its pars pro toto version T019, only a curl) and T- = YSB b’u, depicting a frog(?) with a dotted curl (and its pars pro toto version, T021, only a conch shell) (see Macri & Looper, 2003, pp. 213–214), which are graphically and phonetically similar. They appear to be the product of an adaptive split, and indeed ‘in the first half of the Early Classic period there is generally just one sign’ (Mora-Marín, 2003, p. 226). For Lacadena (2010) b’u is a modification of mu and the script originally transcribed another (Mixe-Zoquean) language lacking the phoneme /b’/, before being adapted to a Mayan tongue. Conversely, for Mora-Marín (2003, p. 203, 205, 207, 226) both signs have Mayan values: mu was acrophonised based on Greater Lowland Mayan *much ‘frog, toad’ and the splitting b’u was also phoneticised from Lowland Mayan *b’u(h)b’ ‘tadpole’.

21 Thus, what to our Western mind is a ‘word’, in Nahuatl is a nuclear clause or sentence-word (Andrews, 2003, p. 6): cf. morpheme a- ‘water’ vs. atl ‘(it is) water’. Daniels claims that in Sumerian, Chinese and Mayan ‘words are generally one syllable long’, and that words in these languages ‘are clearly identifiable’ (Daniels, 1992, p. 83). Yet, it was not words but morphemes that were generally monosyllabic in these languages. In addition, devisers of early scripts seem to have had a notion of morpheme, regardless of whether the target language had mostly free or mostly bound morphemes, i.e. whether the language was isolating or synthetic.

Additional information

Funding

This research is output of the ERC Project “INSCRIBE. Invention of Scripts and Their Beginnings”. The project has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (Grant Agreement No. 771127). The first author would also like to acknowledge the support of the Spanish Ministry of Science, Innovation and Universities under the Juan de la Cierva Formación Post-doctoral Fellowship FJCI-2017-31941.