Creole languages and island vernacular architectures

It is my belief that analogies between Creole linguistic patterns and West Indian vernacular architecture are valid and important. When well constructed, they should open up many important avenues for further research in Caribbean architectural ethnography. They must not be drawn too specifically, however, or they will remain unproductive. Similarities between these two institutions of West Indian culture relate more directly to sociocultural processes than to specific forms. One should begin not only with the forms of the Creole language, but with the dynamic interrelationships between all levels of the post-creole speech continuum. Both architecture and language are forms of social symbolic communication. In both, the adoption of specific forms from a scale of possible alternatives symbolizes one’s identity, values and aspirations.

– Edwards (1983: 194)

Extending feature analysis from creole linguistics to creole architecture

Creole languages are natural languages that develop out of contact between several languages. They are new languages. These ways of speaking have inherited the lexicon of an imperialist nation, often connected to European colonial powers like the Dutch, English, French, Portuguese, and Spanish. Other expanding languages such as Arabic, Japanese, Malay, and the Amerindian language, Tupinamba, have given birth to such new languages. While a creole language inherits the lexicon from the expanding languages, the grammar of creole languages is almost completely innovated and non-inherited. Some traits are from the same language source as the lexicon, and a few can be attributed to other contributing languages. Computational linguists and creolists have measured that up to 80% of the grammatical traits in creoles are innovated.

People in locations where creoles are spoken need houses in which to live. How do people learn to build domestic houses in such environments new and unknown to them? How do these people speak about such new natural and social ecologies and private architectures, especially in small, previously uninhabited island locations? Are there measurable relationships between the development of architectural features like gabled end façades, modular construction, and the use of stone in creole houses and linguistic features like indefinite articles, gender distinctions in personal pronouns, and the occurrence of nominal plural markers in creole languages? Do creole architectures develop in places where creole languages evolve and vice versa? If so, will an empirically based, comparative feature analysis of the domestic house architectures and languages of a set of small islands, using one island, Norfolk Island, South Pacific, as a case study, lead to results relevant to both (creole) architectural history and theory and (creole) linguistics and contact language linguistics? Will such an analysis help bring these fields closer together and move architectural history and theory’s application of both linguistics generally and the architecture is language metaphor specifically out of the nominal and into more empirical domains?

This exploratory article:

1. Identifies what creole architecture is and how it is relevant to the study of contact languages like pidgins, creoles, and dialects, hereafter creoles, and vice versa;
2. Outlines the role of earlier studies in linguistic feature analysis;
3. Uses a previously argued syncretic of nine features typical of creoles and considers these features using data from Norfolk, the Norfolk Island language;
4. Outlines an amalgam of 11 architectural features which are claimed to be typical of domestic creole houses.

I consider a selection of domestic architecture built by the Pitcairn Islanders after 1856 on Norfolk Island and relevant aspects of the language as per the nine typical creole features identified. I have previously claimed Norfolk as an island beach community language and Norfolk Island domestic architecture built by the Pitcairn Islanders as an illustrative example of fusion of creole linguistics and creole architecture. I propose island beach community architecture as a distinct type of island-developed and island-derived domestic creole house, which designates such a creole architectural situation on an island like Norfolk Island. These types of structures are vernacular in their form and creole in their history.

I conclude by arguing that the Norfolk Island creole linguistic and creole architectural situation is representative of English-based island beach community languages in the Pacific like Pitcairn Island English, Bonin Islands English, and Palmerston Island English more generally and the possibility of concomitant island beach community architectures, specifically creole houses, on the respective islands. These buildings and ways of speaking have arisen in non-colonial settings where the architectures and languages are not targeted toward a top-down way of domestic building or vernacular speaking. Rather, built structures and local idioms have developed in locations where architectural and linguistic adaption was essential for the survival of the populations who had arrived on these islands. It is proposed that these houses and individual architectural forms as island beach community architectures likely develop and continue to exist where parallel and associated island beach community languages are spoken.

Creole architecture

Jay D. Edwards, an American anthropologist and architectural historian, completed his PhD in anthropology in 1970 from Tulane University. Born in 1937, Edwards travelled to San Andrés and Providencia Islands in the Caribbean in 1966 to study relationships between English-based creole language forms and social structure (Edwards 1970). Housed within anthropological linguistic theory, this work demonstrates an astute understanding of how intangible culture like language can be compared and contrasted with tangible culture like architecture. Edwards’ principal research field is creole architecture. This subject is founded in his linguistics, e.g. Edwards (1994, 2008). One of his most precise descriptions of the linguistic analogy of creole architecture uses the terms acrolect—the most prestigious dialect or variety of a particular language—basilect— a less prestigious dialect or variety of a particular language—mesolect— an intermediate dialect or variety of a particular language, and sociolect—the dialect of a particular social class. Edwards (1983: 194-195) writes:

In terms of the linguistic analogy, each house may be considered a sentence with its own grammatical structure (in the formal, “structuralist” sense) and its own texture (superficial elaboration, style). Some houses are spoken in basilect, some in mesolect, and some in acrolect. At any single lectal level, houses may be expressed in more or less formal or familiar style. Style ranging applies as well to architecture as it does to language (see, for example, Edwards, Roseberg & Hoy 1976). The rules of stability and change that apply to sociolects in a creole-standard continuum apply equally well to colonial and post-colonial architecture. It is critically important that the cultural system (linguistic or architectonic) be considered in the broadest possible expressions of its diachronic and synchronic dimensions. Portions of this system may be poorly represented on certain islands. Such deficiencies may be overcome by the comparative method through which historical and structural relationships between all parts of the overall system may be explored.

Edwards (1983: 190) claims in a review article of Berthelot and Gaume (1982) that their architectural feature analysis method is the first comparative study of Caribbean (creole) architecture:

Fifteen architectural features are tabulated. They are grouped into sets representing European culture (main entrance on the long side, axial façades, jalousies, and frame construction raised on posts), features of colonization (modular construction, the corner verandah, and mobile huts), the black population (gabled end façades, solid shutters, and gabled roofs), and domestic French architecture (the hip roof, and “shipbuilder’s carpentry”).

These choices of architectural features and their linguistic application are epistemologically, theoretically, and methodologically compatible with the nine typical creole features outlined below. The result comprises a congruent novel architectural feature analysis of this study.

Previous studies in linguistic feature analysis

Since the late nineteenth century, the main theoretical question in creolistics has been to what extent pidgins and creoles are merely reduced forms of another, usually European, language, due to inadequate exposure, or whether they derive grammatically from other, usually non-European, languages re-lexified from and with European vocabulary. From around 1930 to 1980, it was generally accepted that creoles necessarily derive from a pidginised form of a European language. More recently there have been other ideas, including that creoles closely represent the innate human capacity for language. Here, children would have used their bioprogram for language to create creole structures. A somewhat related and currently controversial matter is whether creoles are typologically different from other languages. This typological concern is the focus of the Atlas of Pidgin and Creole Language Structures (APiCS) and the Survey of Pidgin and Creole Languages.

What is essential for exploring how to extend feature analysis from creole linguistics to creole architecture is, first, to identify what linguistic features typical of creoles might be illustrative for island beach community languages. Second, the findings based on this conglomerate of features will then be equated to creole architecture and associated architectural features to ascertain to what extent island beach community languages and their development can be related to island beach community architectures.

Daval-Markussen (2018: 113) writes: “The fact that so many anglicists, creolists and typologists have searched for common creole features strongly suggests that they believe that creoles share a distinctive typological profile – otherwise, none of these authors would have attempted to identify typical creole properties.” He presents a comprehensive survey of some 15 feature analyses of pidgins, creoles, and other contact languages (Daval-Markussen 2018: 87-122). In the chapter titled ‘The search for pan-creole features in the literature’, he breaks up these analyses into the sections: ‘Early lists of recurring creole features’, ‘More recent proposals’, and ‘Recent developments’. These sections culminate in his own proposal for a synthesised feature compilation to determine whether creoles are typologically similar or dissimilar to other languages: ‘A new case study: 9 creole features and 245 languages’. The 245-language feature conglomerate was derived from combining 56 creole languages from the APiCS and 189 languages from the World Atlas of Language Structures (WALS) for which the corresponding features were sufficiently documented, i.e. where the documented language had more than 80% known values. The nine typical creole features proposed in the literature found in the combined APiCS and WALS databases are:

1. Indefinite articles
2. Negative morpheme types
3. Predicative noun phrases
4. Gender distinctions in personal pronouns, but not elsewhere in the language
5. Position of interrogative phrases in content questions
6. Definite articles
7. Order of possessor and possessum
8. Occurrence of nominal plural markers
9. Predicative possession

I consider these nine features by using the Norfolk Island language. According to the APiCS, Norfolk:

1. Has an indefinite article—ar or aa (a/an)—distinct from ‘one’, e.g. Where’s ar house? (Where’s a house?).
2. Has a negative particle preceding the verb, nor, e.g. I nor gut et (I don’t have it).
3. Has predicative noun phrases, namely an obligatory invariant copula, e.g. Bill es bus driver (Bill is a bus driver).
4. Has only third-person singular gender distinctions in personal pronouns, e.g. Bill and Jane es bus driver. He and she es good un, too (Bill and Jane are bus drivers. He and she are a good ones, too).
5. Has interrogative phrases word initially in content questions, e.g. Wathing you doin? (What are you doing?).
6. Has definite articles distinct derived from but distinct from demonstratives, e.g. You bin over dar house? (Have you been over at the house?).
7. Has both possessor-possessum and possessum-possessor word order, e.g. Bill’s hat (Bill’s hat) and Ar hat fer Bill’s (Bill’s hat, literally ‘the hat of Bill’s).
8. Has variable plural marking of all nouns, e.g. I own two house (I own two houses) and Dars one fleas orn yous lieg (That’s one flea on your leg).
9. Has transitive predicative possession rather than the globally more common ‘one child is to/with/at me’, e.g. I gut one little salan (I’ve got one child).

According to these nine features, Norfolk provides illustrative data on varying spectra of a creole feature analysis. I now propose a comparable 11-feature architectural feature analysis of a small corpus of Norfolk Island domestic architecture.

A novel study in architectural feature analysis

1. Main entrance on long side of house
2. Jalousies
3. Frame construction raised on posts
4. Modular construction
5. Corner verandah
6. Gabled end façades
7. Solid shutters
8. Hip roof
9. Shipbuilder’s carpentry
10. Porch
11. Georgian geometry

I consider these 11 features by using architectural data from Norfolk Island. The Australian photographer, Wesley Stacey [1] (1941-2023), created the corpus of architectural photographs taken on Norfolk Island between 1968 and 1972. Examples of Stacey’s photographs of Pitcairn Islander-built creole houses [2] are presented in Figure 1 and Figure 2 (the interested reader is encouraged to consult more examples in the online archive):

1. Both Figure 1 and Figure 2 depict the main entrance on the long side of the house.
2. Jalousies, blinds or shutters made of a row of angled slats, were not common in Pitcairner houses on Norfolk Island. Paned windows of glass, a likely later and more evolved feature than jalousies, are typical in Norfolk Island creole houses.
3. Robert Varman, former Norfolk Island Museum director and archaeologist, describes a house known as Shunna’s such: “Judging by the building materials, this house must be one of the earliest of the Pitcairner homes. Some of the bearers, joists and studs were reused from a convict structure. Some of the doors and windows may also originate from a convict structure. The nails used in the original part of the structure are early in type; square shafted Ewbank’s nails were used and square shafted flooring brads used for the flooring. In some places it appears that even convict nails were reused. The foundations are composed of stone pillars taken from various convict buildings” (Varman 1984: 93). These comments suggest not only that early Pitcairner creole houses reused, refashioned, and adapted available materials from previous convict buildings through what I label was a process of architectural creolisation, but that these earlier houses used existing or moved stone footings and developed frame construction on these footings. Later houses photographed by Stacey feature wooden foundations and early frame-construction features. Varman (1984: 262) confirms that some houses did not use resources from the convict area at Kingston: “As the house was built so far away from Kingston and from convict remains, the foundations are made of piers of natural sectioned Ironwood and Yellowwood.”
4. Varman (1984: 8) hints at the earliest examples of modularity in Pitcairner creole houses: “In the Evans’ style home [a later version of the 1840 domestic houses], a wing was added to the back of the house, resulting in a [sic – ‘an’] asymmetrical plan.” Figure 2 also represents modularity through the gabled addition, likely a later add-on, to the otherwise symmetrical floor plan.
5. Both Figure 1 and Figure 2 present a corner verandah, a feature more developed to the climatic conditions of Norfolk Island and the lifestyle choices of the Pitcairners.
6. Figure 2 illustrates a gabled end façade.
7. Figure 1 and Figure 2 do not have solid shutters or any other type of shutters on windows. The same can be said for all examples in Stacey’s online archive. While not detectable from photographic evidence, archival research suggests internal shutters were uncommon if not completely absent in the stone houses at Kingston, Norfolk Island, and in later domestic creole houses erected elsewhere on the island.
8. Figure 1 presents a hip-roofed cottage surrounded with enclosed verandah in a Pitcairner creole house.
9. The early Pitcairners on Norfolk Island were sea-faring folk and their creole houses exude this fact in thrift and reuse of materials and prudence and utility in design.
10. Figure 1 and Figure 2 exemplify creole houses with porches.
11. Figure 1 portrays Georgian geometry in floor plan. Figure 2 reveals an eccentric interpretation of a largely Georgian-inspired cellular arrangement.

These features and their confirmation-refutation result in the Norfolk Island creole house corpus is perspicacious in terms of how they exemplify architectural change and development across time. Such houses can form the basis of what a feature-based analysis of architectural creolisation might be more generally.

Anarchic tongues, mutinous edifices

I summarise with a different posing of the ‘island vernaculars, island architectures’ and ‘island beach community languages, island beach community architectures’ schemas. I consider languages like pidgins, creoles, and dialects, which are often spoken in marginalised social environments, anarchic tongues. These languages with low speaker numbers force populations to adapt to tricky and unfamiliar natural ecologies such as those found on small islands I have identified. Where a top-down approach often exists in coloniser-imposed language situations where speakers’ linguistic human rights are compromised, flatter hierarchies can also arise. These anarchic tongues, or outcomes of the decentralising of language and social power, are linguistic results worthy of scientific attention. And, as I have hinted throughout, the architectural results of these people speaking such anarchic tongues are intertwined with language and cultural development in place. I term these buildings mutinous edifices. They arise in flat linguistic and architectural hierarchies; they evolve in a rebellious manner out of the arcane need for humans to be housed; they are mutinous because they resist classification into particular types in the same way the languages in which they develop in parallel withstand uncomplicated categorisation; and, in the case of Norfolk Island, they are built by descendants of mutineers.

The three posings with which I conclude are:

1. Island vernaculars, island architectures
2. Island beach community languages, island beach community architectures
3. Anarchic tongues, mutinous edifices

The method, theory, and findings presented in this article are part of a larger research project which tests theories analogising linguistic creolisation with architectural creolisation in small island environments. This research project is based in the following definition of creole architecture:

Creole architecture and the process of architectural creolisation are what take place when a group of people, especially those who themselves are culturally mixed, inherit often due to relocation, a pre-determined and already-built building style as an architectural pidgin. When this pidgin architecture becomes vernacularised through necessary building and cultural adaption and expanded in terms of its meaning (semantics) and size and ordering (syntax), it becomes creole architecture. Further, in locations where there is parallel language development taking place in the architecture makers’ society, it is proposed that the changes in built environment may have parallel linguistic changes, because of the nature of the evolving familial structures and language change. That is, linguistic creolisation appears either to have begotten or have occurred at a similar time as architectural creolisation and vice versa. (Nash 2024: 80)

It is hoped that what I have presented here may offer architectural historians different ways of thinking about cultural change in architectural vernaculars across generations so that linguists and architectural historians might be able to work together more effectively in future. This correspondence should be productive in both directions; creole architecture reasoning could offer much to creolists and linguists working in comparable language-based environments. These exploratory and hypothetical frameworks need to be further tested from the standpoint of both architectural history and theory and linguistics. Such studies would ideally be comparative studies involving larger samples through time of both architectural and linguistic documentation to ascertain drivers of creolisation and change. Although I have suggested that the island beach community languages and parallel island beach community architectures of islands like Pitcairn Island, Bonin Islands, and Palmerston Island would likely be illustrative, research in case studies beyond the Pacific and the Caribbean would also be welcome.

References

Berthelot, J. and Gaume, M. 1982. Kaz Antiye: jan moun ka rete (Caribbean popular dwelling/L’habitat populaire aux Antilles) [translators: K. Bowie (English), R. Fontès, J-P. and J. Sainton (Creole)]. Paris: Editions Caribéennes.

Edwards, J.D. 1970. ‘Social linguistics on San Andrés and Providencia Islands, Colombia’, PhD thesis, Department of Anthropology, Tulane University, New Orleans, Louisiana, USA.

Edwards, J.D. 1983. ‘The first comparative studies of Caribbean architecture’, New West Indian Guide / Nieuwe West-Indische Gids 57(3/4): 173-200.

Edwards, J.D. 1994. ‘The Origins of Creole Architecture’, Winterthur Portfolio 29(2/3): 155-189.

Edwards, J.D. 2008. ‘Unheralded contributions across the Atlantic world’, Atlantic Studies 5(2): 161-201.

Edwards, J.D., Roseberg, M. & Hoy, L.P. 1976. ‘Conversation in a West Indian taxi’. Language in Society 4(3): 295–321.

Michaelis, S.M., Maurer, P., Haspelmath, M. & Huber, M. (eds.). 2013. The Atlas of Pidgin and Creole Language Structures. Oxford: Oxford University Press.

Nash, J. 2024. ‘Linguistics and architecture, creolistics and history, or, is Norfolk Island architecture (a) creole?’, In P. Memmott, J. Ting, T. O’Rourke & M. Vellinga (eds.), Design and the Vernacular: Interpretations for contemporary architectural practice and theory. London: Bloomsbury: 77-95.

Varman, R.V.J. 1984. Survey Study of the First, Second and Third Settlements on Norfolk Island. Canberra: Australian Heritage Commission.

Notes

[1] Stacey is a well-reputed photographer of Australian colonial architecture accredited with more than 3,000 photographs taken in the late 1960s and early 1970s covering vernacular timber constructions, Georgian structures, homestead buildings, and historic towns.

[2] I use the expressions ‘Norfolk Island creole house’ and ‘Pitcairner creole house’ interchangeably.

Cover photo: Church, Palmerston Island, South Pacific, source: http://lh3.ggpht.com/_i0O0FDaxgcA/THsN6HxXtmI/AAAAAAAAE9A/DZB4-ZVObyk/s1600-h/P80702912.jpg

Joshua Nash is the editor of Some Islands (someislands.com).

Island languages Global Languages Day, Sept. 17, Dokk1, Aarhus, Denmark A creole book and a dedication to a creolist: Philip Baker Book review: An extinct creole language of the Danish West Indies Derek Bickerton (1926-2018), the insular linguist, and his work Documenting fishing records and wet placenames around Dudley Peninsula, Kangaroo Island, South Australia