NEW PERSPECTIVES ON JAMAICAN ARCHAEOLOGY ROBERT R. HOWARD

only the most cursoryarchaeologicalinvestigation. Recent investigationsstill yield no sign of preceramic Since Meso-Indiansiteshave not been found, or of Arawak occupation of Jamaicabeforeearly Period [II in the Rouse chronology. Jamaican sites indicate a we mustturnto the evidencefo,r Arawak occucharacteristicallyArawakan culture, but they lack the patioin.The earliestknownArawak site in Jaceremonialelaborationdeveloped in Haiti. Jamaicanpot, on the northteryshows relativelylittlechange throughtime and only maica continuesto be LittleRiver laterArawak Unlike island. of the coast central present: are styles minor local variation. Four distinct on thecoasltand has (1) Little River fromearly Period III; (2) White Marl, sites,LittleRiveris directly the dominantstyleof the island; (3) Montego Bay, prob- been almostcompletely erodedawayby thesea. Red Ware ably the latest in time; and (4) a little-known no radiocarbonsamples that unfortunate It is closest whose chronological position is no,tclear. The because it now apthe site from rescued were are with Bani of Cuba and Meillac of stylisticaffiliations Haiti. The late Carrier style of Haiti did not diffuseto pears doubtfulif enough materialremainsfor Jamaica. Certain distinctiveburial practicesand features this to be accomplished. The date for Little of village construction have been revealed by recent Rivertherefore mustat presentbe derivedfrom excavationsat White Marl and othersoutherncoast sites. and associateddates with ceramiccross-checks in the West Indies. As Dematerial elsewhere wasthelastofthemajorislandsof JAMAICA of thesite,has pointedout, the discoverer Wolf, the Antillesto receivearchaeologicalattenan extensionof to represent seems the po,ttery last in the done what has been tion. Much of of Puerto Rico tradition Cuevas-Ostiones the decade has been, of necessity,surveyworkand places Little This 1953). into (DeWolf Jamaica testexcavationsratherthan intensiveinvestigaIII in Rouse's Period part of early in the River tion of particularsites. Only at the important a date of and suggests 1951) scale (Rouse time coast the southern on villagesiteofWhite Marl has a major excavationcampaignbeen begun, around A.D. 500 for initial occupation of the stages. There- island. and thisis stillin its preliminary No other sites yieldingmaterial similar to fore many of the conclusionsreached in this River are knownin Jamaica,althougha Little summaryare tentativeand will undoubtedlybe in the refuseremovedfromthe hisfew sherds becomes modifiedas more complete evidence toric of Sevilla Nueva by C. S. Cotterapsite availablein thenextfewyears. to the LittleRiver style. Since to belong pear At present,the Sub-Tainobranchof theArathis earliestSpanish settlementis only a short wak appears to representthe only pre-Columdistance fromLittle River and is also on the bian inhabitantsof Jamaica. Ripley P. Bullen coastalplain near thesea, it seemsprobablethat and I and several competentamateursin Ja- theremay have been an earlyArawak village maica have searchedforevidenceofpre-agricul- at thissite which had been abandoned centurtural Meso-Indianculturesthat correspondto ies beforethe Spaniardsbuilt there. The lack the Ciboney of Cuba, but no such sites have of materialbelongingto LittleRiverstyleelsebeen discovered. If we considerthe fact that whereon the islandsuggeists thatthe firstoccuMeso-Indianoccupationshave now been coin- pation of Jamaicawas small in scale and was firmedforall the major islandsof the Greater confinedlargelyto the northernco,ast. Antilles except Jamaica,it seems most likely Just when the majo,rArawak migrationto that such sites will eventuallybe discovered Jamaicabegan and which areas were firstoccuthere as well. Certainlythe archaeologypro- pied cannotbe said forcertainat present.The vides no obviousexplanationforthe bypassing earliestradiocarbondates we have forJamaica of Jamaicaby Meso-Indianmigrants.It should come from the White Marl site midway bebe rememberedthat certain portionsof the tweenKingstonand Spanishtownon the southJamaican coastal area, such as the Portland ern coast. These dates, which were obtained Ridge region and much of the western and fromcharcoal in the bottomlevels of occupasouthwesternshoreline,which representthe tion,clusteraroundA.D. 900. The mainArawak typeof ecologicalsettingmostfavoredby Meso- occupationmaygo back a centuryor twoearlier Indianveorlesin Cuba and Haiti,have received but probablynot much beforethat. ABSTRACT

250

HOWARD

I

JAMAICAN ARCHAEOLOGY

251

wereprobJamaicanArawak sites fall into threemain the villages,defensiveconsiderations categories:(1) middendepositsat villagesites, ably secondary.The only considerablearea of As is character- interioroccupation seems to, have been the (2) caves,and (3) petroglyphs. sectorof the island. Here, as one isticof other Sub-Taino remainsin the West north-central Indies,thereis an absenceof plazas,ball courts, movesawayfromthesea towardthemountains, or otherceremonialstructures.Caves, which land shells replace marine formsin the midwas probablymoreimoccurin largenumberin Jamaica,seem to have dens.Manioc agriculture been used almostentirelyforburial of persons portantforthesevillagersthan it was forthose of highrankand forrelatedreligiousactivities. nearer the coast, althoughmost midden sites A few scatteredpotsherdsand shells in some are near land that could have been used for of the caves indicateoccasional use as tempo- cultivation. Most Jamaicanmiddensare shallow,varying for huntersor gatherersof shellrarysheliters fish.No new caves have been discoveredin the froma few inches to,3 ft. in depth. A few excavation exceptional si(tes,such as White Marl and last decade, and no reallysystematic has everbeen undertakenat any of the known Tower Hill on the southerncoast,show refuse accumulationsof 6 ft.or moreand mustreprecave sites. continuous Of the nine reportedpetroglyphsitesin Ja- sentseveralcenturiesofmore-or-less maica,severalhave been seriouslydamaged by occupation. At White Marl, forexample, the most- radiocarbondates indicateat least 400 yearsof vandalsin recentyears. The petroglyphs, use of thesame location. No eviand a few uninterrupted representations ly anthropomorphic animals, are very crude and have facial and dence has as yet been uncoveredof any condelineatedin the mostrudimen- sistentpatternof arrangementat any of these bodyfea!tures locationof most petroglyph village sites,except forthe casual clusteringof The fashion. tary sites away from areas of occupation and in dwellingsaround a central open area. Most ofaccesssuggeststheymayhave Jamaicanvillagesdo not appear to,have been regionsdifficult servedas shrineswheresome particularclass of occupied forverylong periods,but where the paittern can somerefuseis deep an interesting zemiswas propitiated. themostcom- timesbe observedin the deposition.At interVillage middensitesconstitu,te mon categoryof Arawak remainsin Jamaica. vals 2 to 3 ft. apart, a layer of sterile,white During the last decade the number of such limestonemarl 1 to 2 in. thickhas been laid known sites has been nearlydoubled, thanks down. This marlingis clearlyartificialand nolt largelyto the systematiceffortof severalwell- the result of wind action. Its purpose is not qualifiedamateurs: Dr. JamesLee, J.Tyndale- clear; it may have servedto "sweeten"the site Biscoe,FatherF. J.Osborne,and C. S. Cotter. periodically,or some patternof ceremonialreThese men have coupled zeal with the unique newal maybe indicated. The presenceof burialsin middens,hitherto opportunitiesprovidedthem in the course of theirroutineactivitiesto,recordaccuratelythe unknownin Jamaica,has been revealed in the village last fiveyearsby testexcavationsin threeof the locationsof many hi,therto-unreported sites.For themostpart,onlysurfacecollections deepest middens at the White Marl site. A have been made, but the materialsthus accu- single burial was discoveredat the bottomof mulatedhave widened measurablyour knowl- each midden.All were flexed,twoof thebodies beinginterredon the side,the thirdin a seated edgeofJamaicanpottery. The nearly200 known village sites confirm position. Each burial appeared to,have been patternI notedin mysurveyof placed on the originalgroundsurfaceand the the distribution 1947-48 (Howard 1956). Most villagesare on refuse then deposited on it over a period of hilltopsoverlookingthe coastal plain and with- time. None of the burialsseems to have been in reasonablyeasy access to the sea. The im- intrusive.Burial accompanimentwas entirely portanceof shellfishin the Arawak diet, as lacking. These findspresentseveralpuzzlingfeatures. attestedby enormousquantitiesof shells in the absence of grave goods and the generally the to The proximity the refuseheaps, accountsfor nature of the intermentssuggestthat locations elevated humble for the and preference sea, oversiteson the coastal plain can easilybe un- these were hardly personsof rank. The one derstoodby anyonewho has lived in Jamaica. skeletonfully analyzed thus far is that of a in any of youngman in his late twenties.While it is posAs thereis no evidenceof fortification

252

AMERICAN ANTIQUITY

[ VoL. 31, No. 2, 1965

sible that additionalburialswill appear when thatitsdevelopmentcan be tracedovera period the refuseheaps are entirelyclearedaway,I am of severalcenturies.Most of the aboriginalpotinclined to, doubt that any great number of tery found in Jamaica belongs to this major in skeletonswill be found under any singlemid- style,and thereis a remarkablehomogeneity and rangeofdecoraden. This seemsto representsome specialburial techniquesof manufacture than a prevailingmode of inter- tive modes whereverthe style occurs on the patternraltiher ment formostof the villagepopulation. Only island. Some regionalvariationsundoubtedly additional detailed excavationof several mid- were present,but these appear to have been minorand unimportant.MostWhite Marl vesdens can clarifythis.point. In 1963, a considerablearea in the centerof sels fall into two basic shape categories: (1) the White Marl village was cleared and ex- round and (2) boat-shaped.Most potsof each amined for posthole remainsin the hope of shape are shouldered and round-bottomed. on size, mode of construc- Round bowls varyin size from12 to 20 cm. in gaininginformation tion, and arrangementof Jamaican Arawak width and froim8 to 15 cm. in depth. Boatdwellings.,but no conclusive results were shaped vesselsaverageabout 15 cm. in length and 10 cm. in width. The largestknown,which obtained. While a varietyof stone,shell,and occasion- is froma burial cave on the northerncoast, have been foundin Jamaican measured 40 cm. in length,28 cm. in depth, allywood artifacts archaeologicalsites,potteryis by far the com- and 25 cm. in width. Complete vessels are monestmaterialencountered.Althougha pre- known almost entirelyfromburial caves; they liminarydescriptionof Jamaican poitteryhas are rarelyfound in middens. Compartmented been givenelsewhere(Howard 1956), it seems vessels,miniaturepots, and water jars are the to summarizeitssalientcharacteristics onlyothervesseltypes.The firsttwo,unknown pertinent once more,withcertainadditionsand modifica- in Jamaicauntil recently,have been found in middensand at White Marl. Waiter tions made possible by the last five years of north-coast excavation and collection on the island. At jars (Fig. 1 i) were apparentlymorenumerous, thereappear to,be threedistinctlocal althoughthey seem to be lackingfrommany pres.ent, whichmayor maynotbe of areas altogether.They are usuallyrepresented styles,and a fourtlh Jamaicanorigin.The earliestof thesein timeis onlyby the veryheavyarchedhandlesand, un, LittleRiver,firstdescribedby DeWolf in 1953. til a fortunatefind was recentlymade by Dr. This style,as notedabove,seemsto representan JamesLee, we did no!tknow the formof the extensioninto,Jamaicaof the Cuevas-Ostiones vesselto whichsuch handleswereattached. Aside fromthe water jars, JamaicanWhite potteryof PuertoRico. DeWolf feelsthat it is closer to earlyOstiones than to eitiherCuevas Marl potteryis relativelythin-walledand unior late Ostiones. LittleRivershowsa highinci- formin thickness,the average thicknessbeing dence of D-shaped handles,loop handles, flat- about 8 mm. The vessels are coil-constructed, boittomedvessels,necked vessels,vessel walls and thereis littleindicationof added coloring with outcurvingrims,extensiveuse of paint, mattero,rartificialtemper. The clay of fired and paucityof incised and affixeddecoration. vessels varies in co,lorfrombrickred through All of thesetraitsare conspicuousby theirrarity reddishgray,brown,and black,withyellowand or absencein laterJamaicanpottery.Certainly buffprevailingin some partsof the island. No Little River does not seem to have influenced slip was applied,but mostpoitsreceiveda dull the majorJamaicanpotterystyleto any marked polish. On the whole,the potteryis well-made if somewhatuninspiredfroman and well-fired, degree. This second style (Fig. 1 d-j) I have tenta- estheticpointofview. tivelycalled JamaicanMeillac because of its Comp:aredwithMeillac or Bani, White Marl withthatstylein Haiti. There are shows lesis decoration and fewer decorative close affinity also many traitsshared with Cub;anBani pot- modes. At least 50% of the potteryexamined tery. I feel thatthe timehas now come to give has no decorationwhatsoever.Paintingis largeitmore ly absent. The usual techniquesare incision, thestylea name ofitsown to distinguish clearlyfromits Haitian and Cuban relatives, application,punctation,and modeling. Vessel and I proposethe termWhite Marl, since it is shoulders,handles, and lugs were the main at this site that the style appears in its most areas of decoration. Incised designsare invarithe mostcommonbeinggroups and completeformand it is here ably geometric, characteristic

HOWARD

]

Fiu.

JAMAICAN ARCHAEOLOGY

1. Sherds ofJamaican pottery.

.-C,MontegoBay style;

of alternating,obliquely parallel lines applied to the shoulder. Crosshatching, so commonin Meillac,is quite rarein Jamaica(Fig. 1 c). The favoritedecorationforrimtopsis thatof closely spaced parallel incisions,usually about 5 mm. apart,which are sometimescut deeply enough to givetheedge a serratedappearance (Fig. 1 c). Curved-lineincisionoccurson veryfewsherds. Punctationusuallyaccompaniesline incisionas

cl-i,

253

White M.arl style;k-n, Red Ware.

a decorativemode. Evenlyspaced parallel rows of do!tsare commonlyplaced on theshoulderor on the rimand above any inciseddesignwhich may be present. The techniqueof application employedthanthatofincision. is less frequently The mostcommonis a serrateddouble-curved motif. Another is a limb designin which the applied stripis flattenedand broadenedat one end and marked with two verticallyincised

254

AMERICAN ANTIQUITY

[ VOL.31,

No.2, 1965

Two otherpotterystylesremainto be noted. designsthat vaguelyresemblethe limbsof an animal or reptile. Such designs are usually The first(Fig. 1 a-c) I have called MontegoBay found on vessel shoulders. Lugs and handles since it was firstdiscoveredat the Fairfieldsite show the greatestvariationand are the most nearMontegoBay and seemstobe concentrated decoratedfeaturesof Jamaicanpot- on the northerncoast.It alsooccurssporadically consistently quantery.Fourmain typesofhandlesare found: (1) elsewherein Jamaicabutnotinsignificant a rare loo,phandle that probablyreflectsLittle tity. Montego Bay is characterizedby much River influence;(2) a flaringtipped handle thickervessel walls and probablymuch larger (Fig. 1 d), whichis quite common;(3) a cylin- vessels,althoughno completespecimenshave notablefor drical knoblikehandle; and (4) a distinctive been found.The styleis particularly arched handle encounteredon water jars (Fig. iltsuse ofincision.The favoritedesignconsistsof parallellines(about 1 i). Handles of the firstthreecategorieswere groupsof short,horizontally usuallyplaced at each end of boat-shapedves- 1 cm. apart) thatare heavilyincisedaroundthe decoratedwithincised vesseljustbelow the rim (Fig. 1 c). Short,wide, selsand are oftenfurther designs. Very common, especially on round obliquelyp,arallellines formanothercommon bowls, are wedge-shapedlugs, almo,stalways design (Fig. 1 a). Virtuallyall the incisionon verticallyincised,which are placed in opposing Montego Bay potteryis much deeper, bolder, pairs at the vessel inturn,parallel to the rim and less carefullyappliedthanthatfoundin the of theshoulderat thebiase White Marl style. This typeolfincisionseems (Fig. 1 j). Perforation of the handles is sometimesfoundon Jamaican to be in keepingwith the coarse,heavy nature of the potteryitself.Applicationis oftencompottery,but it occurs infrequenitly. bined with incision to produce zoomorphic used Modeling does not seem to have been forms (Fig. 1 b). The temporalpositionofMonit When pottery. oftenin decoratingJamaican tego Blay and its relationshipto White Marl or lugs modeled of form takes the it occur, does handles which may be eitheranthropomorphic remainto be determined,but I am inclinedto a laterperiodstyle. A few can be believe thatit represents or zoomorphicrepresentations. A final styleis Red Ware (Fig. 1 k, m), a few particularlysnake and turtleheads, identified, and one quite lifelikemonkeyhead was found sherds of which have been found in widely recentlyon the northerncoast (Fig. 1 g). One separated parts of Jamaica but never in any hintson concentration.The potteryseems verysimilar or two,human heads give interesting to the Red Ware reportedby Krieger(1931) Jamaican of profiles Rim 1 f). (Fig. hair styles potsare of threetypes: (1) rounded,(2) taper- fromthe Dominican Republic and by Fewkes ing,and (3) rectangular.Of these,therounded (1907) fromPuerto Rico, where it occurs in rim is by far the most common. Fillets were fairlyearlylevels in bothcases. The distinctive frequentlyapplied along the outer surfaceof color of the potterywas achievedby the applithe rim,and sometimesthe rimcoil was thick- cationof a red slip whichwas thenhighlypolened (Fig. 1 h). Ridgeson theinsideof the rim ished. In this ware modelingis accomplished throughextensionof thevesselwall, and affixaare rarelyfound. often tionis absent. Simpleincisionis added to delinobjects of pottery classes Two o,ther foundin Jamaicashouldbe no,ted.Griddlefrag- eate human and animal facial features (Fig. ments,averagingbetween2 and 3 cm. in thick- 1 l). Ribbon-loophandles are commonin the ness,comefromround,beveledgriddlesbetween Red Ware (Fig. 1 m). It remains to be deter30 and 35 cm. in diameter. These griddles mined whetherthe Red Ware findsare trade are poorly fired and are never found intact sherds or whetherthey representa briefand occupationof Jamaica in middenrefuse.They wereundoubtedlyused numericallyinsignificant forcookingmanioc cakes, and theirfrequency comparableto that of Little River. The eviatteststo the importanceof manioc cultivation dence seems to point to a relativelyearlydate pattern.Conical potteryob- forRed Ware. in the subsisitence jects,usuallytermedpestlesby local collectors, The Carrierstyleand relatedceramicdevelare probablyamulets,since most of the speci- opments of Period IV in Hispaniola and the mens are perforatedat the small end for sus- rest of the GreaterAntilles never reached Japension and several are crudelyincised with maica. The essentialconservatismand resistance to outsideinfluencerevealed in Jamaican human facialfeatures.

H4OWARD

1

JAMAICAN ARCHAEOLOGY

potteryas a whole is onlyone of manyindicawhichis confirmed tionsof culturalretardation, onlytoo clearlybytherestof the archaeological remains.Whetherfuturearchaeologycan give us the answersto thispuzzlingproblemin culturaldynamicsremainsto be seen. DEWOLF, MARIANW.

1953 Excavation in Jamaica. American Antiquity, Vol. 18, No. 3, pp. 230-8. Salt Lake City. FEWKES,J.W. 1907 The Aboriginesof Porto Rico and Neighboring Islands. 25th Annual Report of the Bureau of Ethnology,pp. 1-220. Washington.

255

R. 1956 The Archaeology of Jamaica: A Preliminary Survey. AmericanAntiquity,Vol. 22, No. 1, pp. 45-59. Salt Lake City.

HOWARD, ROBERT

H. W. 1931 Aboriginal Indian Potteryof the Dominican Republic. United States National Museum, Bulletin 156. Washington.

KRIEGER,

ROUSE, IRVING

1951 Areas and Period of Culture in the Greater Antilles. SouthwesternJournalof Anthropology, Vol. 7, No. 3, pp. 248-65. Albuquerque. UNIVERSITY OF WISCONSIN

Milwaukee,Wisconsin June,1964