Atlatls to Bows: More on the Skagit River Atlatl

Starbucks, New Brunswick, New Jersey

My post on the atlatl found at the mouth of the Skagit River north of Seattle seems to have led one reader to ask about it in a forum for modern atlatl makers and users.  The responses are interesting.  One respondent linked to an article from the 1960s with more detailed information which is available free online.  This article, by Charles Borden, has some very good pictures of the atlatl, which was at some point acquired by the University of British Columbia Museum of Anthropology, which has even better pictures and more information on its website.  Borden’s analysis focuses mainly on the iconography of the elaborately carved figure, which he puts in the context of ethnographically known imagery from various Northwest Coast cultures representing sea monsters and other mythological creatures with similar characteristics to the one on the atlatl.  He argues, not entirely convincingly, that it represents an early form of the important creature known as the Sisiutl, which is usually represented as a two-headed snake but which can take on other forms as well.  Whether or not he is right about that particular identification, Borden does make a convincing case that the atlatl fits easily into the artistic traditions of the Northwest rather than being an import from elsewhere.  He also argues that it is likely very old, and tentatively suggests that it may be contemporaneous with the Locarno Beach site in Vancouver, which produced an atlatl hook made of antler.  The Locarno Beach site defined the Locarno Beach Phase, which now seems to be dated to around 3500 to 2500 radiocarbon years before present.  As I mentioned in the previous post, the Skagit River atlatl was apparently later radiocarbon dated directly and assigned to the Marpole Phase, which dates to around 2000 to 1500 radiocarbon years before present.  (According to the UBC Museum website the exact date was around 200 AD.)  Borden was therefore off by quite a bit in suggesting that the atlatl was contemporaneous with Locarno Beach, but of course he had less information to go by than is available now.

Also, John Palter recently commented on a post in which I discussed an article of his on atlatl weights, pointing to a more recent article in which he bolsters his theory that they were associated with flexible atlatls by discussing the attitudes of modern atlatl users toward the advantages of flexible, weighted atlatls over more rigid types.  As with the forum discussion on the Skagit River atlatl, this shows the interesting insights on atlatl use that can come from the large corps of amateur atlatl users and their extensive experimental experience with atlatls.  This is a very different approach to learning about atlatls than the abstract study of surviving ancient specimens more typical of archaeologists, and I think the two approaches used together can be quite complementary. I’ve mentioned this issue before with regard to interpretation of an atlatl petroglyph.

Read Full Post »

Atlatls to Bows: A Suspiciously Large Arrow

Big Arrow, Albuquerque, New Mexico

In comments to the previous post, pato’ links to a recent press release on the discovery of an atlatl dart in a melting ice patch near Yellowstone.  This type of discovery is becoming more common as global warming causes ice patches and glaciers to melt at an unprecedented rate, releasing artifacts that have been frozen in them for centuries or millennia.  Because of the protective qualities of the ice, artifacts made of perishable materials like wood like atlatls and darts often survive there.  These discoveries are particularly helpful for the study of hunting paraphernalia, because many of the artifacts in the ice got there when they were lost during hunting expeditions in harsh territory that would have been unsuitable for permanent or temporary habitation.  Quentin Mackie has a good discussion of this find and ice patch archaeology in general in which he links to an earlier post of his discussing a much earlier find from British Columbia.

That find was a projectile, made of wood with attached stone point, found by a surveyor near Tsitsutl Peak in 1924.  It was mentioned briefly in news reports at the time, then disappeared into the collections of the Royal British Columbia Museum until it was rediscovered by a curator there who did some testing of it that was published in 2005.  Radiocarbon dating of the wood came up with a 2-sigma calibrated range of AD 1482 to 1639.  An unsuccessful attempt was made to date a sample of the sinew used to haft the point to the shaft as well.  This is quite late, and it suggest that the weapon is an arrow rather than an atlatl dart, as the bow and arrow would certainly have been introduced by that time and the general thinking in this region is that the atlatl was no longer in use then.

Closeup of Big Arrow, Albuquerque, New Mexico

The problem with this conclusion in this case, however, is that the thing is huge.  Including the point, it’s 89.5 cm long and broken at the end, indicating that it was originally even longer.  The shaft alone (excluding the point) is 86.2 cm long.  Recent ice patch finds from the Yukon, discussed by the authors of the paper as a comparative context for the Tsitsutl artifact, show that late prehistoric arrows have shafts ranging in length from 52 to 73 cm, all much smaller.  Atlatl darts from the Yukon are larger, and some are around the same size as the Tsitstul artifact, which the authors estimate to have probably been about 120 cm long originally.  Comparison to some other methods for determining if a point came from an arrow or a dart reveals that the Tsitsutl point doesn’t exactly match either the dart or arrow point characteristics, but in most size attributes it is closer to the dart side.  There is also some ethnographic and ethnohistoric evidence of atlatl use in the contact period on the northern Northwest coast, including an eyewitness account of atlatl use in maritime hunting among the Tlingit in 1788 and a few Tlingit atlatls from the eighteenth and nineteenth centuries in museum ethnographic collections.  Tsitsutl Peak is not in ethnographic Tlingit territory, but it isn’t very far away and certainly part of the same general cultural area.  The Tsitsutl artifact is also made of yew, which was not a common type of wood used for arrows ethnographically.

Now, I look at that evidence and conclude that this is probably an atlatl dart, and that it therefore suggests that the atlatl was in use in this area for longer than archaeologists have generally thought.  Perhaps, as in the Arctic, the atlatl was used primarily for maritime hunting while the bow and arrow was used on land, although the fact that this artifact came from an inland setting is problematic for that theory.  Oddly, however, the authors of the paper analyzing it conclude that it must be considered an arrow, apparently based almost entirely on the late date and a supposed resemblance in form to arrows and lack of resemblance to known atlatl darts (which are of course much older).  They even say that the late date on the wood is so reliable that it doesn’t matter that the attempt to date the sinew failed!  In this circumstance, where the dating doesn’t clearly match the expected form of the artifact, I would say that it would be particularly important to get as many dates as possible.  They don’t really explain why they don’t find the similarity in size to known atlatl darts and the ethnohistoric evidence convincing.  Certainly there are problems with classifying weapons as darts or arrows based solely on size, an issue that I’ll discuss in a later post, but the size difference between the two is pretty well established ethnographically and archaeologically and I think they should have at least acknowledged that and explained their reasoning in making their conclusion contrary to that line of evidence.

While some aspects of the paper’s discussion strike me as dubious, this is an important find, and it’s good to see it published after languishing in obscurity for so long.  Ice patch archaeology has a lot of potential for improving our understanding of the past, but that potential can only be realized if the stuff is collected, studied, and published.

Keddie, G., & Nelson, E. (2005). An Arrow from the Tsitsutl Glacier, British Columbia Canadian Journal of Archaeology, 29, 113-123

Read Full Post »

Atlatls to Bows: Reader Response

Atlatl Petroglyph, Atlatl Rock, Valley of Fire State Park

Jim Weller, who has extensive experience with atlatl construction and use, e-mails with some very interesting thoughts.  About the atlatl petroglyph on Atlatl Rock at Valley of Fire State Park in Nevada, which I have used to illustrate some earlier posts, he says:

What I find very interesting . . . is that the atlatl in that petroglyph doesn’t have a hook.  It’s forked at the end, which I assume means it had a string across the fork and the darts had nocks like an arrow.

. . .

This page here is what made me think the atlatl in the petroglyph was fork-and-string.  The business end of the 2nd atlatl shown looks just like the petroglyph, although it lacks finger loops on the other end. http://www.primitiveways.com/loop_cord_atlatl.html

Also, on finger loops:

As to finger loops, in my own tinkering with atlatls, I’ve developed a preference for them.  This is because the human wrist moves a lot more up-and-down than it does left-to-right, and the more wrist movement you can put into the throw, the faster the dart goes.  Without finger loops, throwing is like chopping with a hatchet (or throwing a javelin without an atlatl), limiting you to the left-right wrist motion. With finger loops OTOH, throwing is much like with a baseball, allowing the full up-down wrist snap towards the end of the throwing motion.

I think there’s enough difference in power (and accuracy) with finger loops for them to have been pretty much standard equipment.  Sure, you don’t absolutely need them, and there are many folks today who don’t use them and still achieve excellent results.  However, if my life depended on an atlatl, I’d definitely use them.  They have physics on their side.  And there seem to be more examples of atlatls from all over the world with some sort of finger loop than there are without.  Leather straps, shell crescents, and holes drilled through wide handles.  Thus, I’m not surprised that the loops are emphasized in art.  I prefer the leather straps because they don’t chafe and pinch my fingers as much as the hard types of loop.

. . .

Note in the petroglyph that the atlatl handle is much narrower at the loops than elsewhere.  This is a pretty necessary design feature for using loops on an atlatl that’s more than about 1/2″ in diameter.  So, if you find just the stick and wonder if it ever had loops, look for this narrow place if the rest of the handle is wider than 1/2″.  If there’s no narrow place, then it almost certainly never had loops.  OTOH, if the whole thing is only 1/2″ wide, then there’s no good way to tell.

The reason you need the narrow place on wide handles is because when using loops, the atlatl goes between the index and middle fingers, which still have to wrap back around on top to hold the dart.  The narrow gap between these fingers and their lack of opposability puts a limit on how big an object will fit between them comfortably.  If the atlatl is more than about 1/2″ wide there, you CAN use it, but each throw hurts the inside of your index finger just below the 1st knuckle, and you have to tense up the whole hand and wrist to get a good grip on the dart, which decreases power and accuracy.  Not fun.  But OTOH, if the whole handle is that narrow, you have to tense up your hand anyway to hold it with your thumb and other fingers below the loops.  So the best design is wide enough at the butt for a relaxed grip for the thumb, pinky, and ring fingers, and narrow above for a relaxed grip with the index and middle fingers.

To see how this works, make a peace sign keeping the thumb, pinky, and ring fingers in a comfortable circle so none of them touch the palm.  That’s a good diameter for the lower part of the handle (or the whole handle if you’re not using loops).  Now, keeping your middle and index fingers as far apart as possible at their 1st knuckles, bend them down and in until their tips touch each other and the tip of the middle finger is touching the end of the ring finger.  Look how much smaller the gap between the index and middle finger knuckles is compared to the gap between the other fingers and the palm.

Also, via the interesting Blackwater Draw blog, I see that John Whittaker of Grinnell College has an extensive annotated bibliography on atlatls available on his website.  From it I see that he disagrees with both Calvin Howard and Bob Perkins about the physics of the atlatl.  I don’t know enough about the thing to judge who’s right.  Gaining that kind of knowledge really requires substantial personal experimentation, and while I could do that, I’m mostly interested in the cultural and historical implications of these technologies rather than the details of their operation.  Still, there’s a ton of literature out there for anyone who is interested, and Whittaker’s bibliography would be a great place to start.

Read Full Post »

Atlatls to Bows: A Serendipitous Lion Hunt

Very Large Array, Plains of San Agustin, New Mexico

Like atlatls, but to an even greater degree, bows are rare in the archaeological record because they were made of perishable materials.  While some types of atlatls had more durable attachments such as hooks and weights, bows were almost always made of wood and various fibrous materials, except in some areas where they were made of horn or antler.  Bows are thus exceedingly rare in the archaeological record, and when they do appear it is usually just as fragments.  Many museums have large collections of complete bows, but these come almost entirely from modern ethnographic collections and are not necessarily the same types that were used in antiquity.  Indeed, there seems to have been a major change in bow technology in the late prehistoric period throughout North America, in which the simple self bow was replaced by a more elaborate sinew-backed recurved type which was both smaller and more powerful.  Ethnographic examples are almost always recurved, so understanding the older self bows requires study of the few archaeological examples available.  These survive only under conditions of exceptional preservation, such as in caves and rockshelters.

One important discovery came about rather accidentally.  In the 1930s, the University of New Mexico did some archaeological surveys in the southwestern part of the state, particularly in the relatively unexplored area in and around the Gila Mountains, between the Plains of San Agustin to the north and the well-known Mimbres Valley, famous for its black-on-white pottery, to the south.  During one of these surveys, near Gila Cliff Dwellings National Monument, Frank Hibben, a graduate student at UNM who would go on to have a long career there as a professor of anthropology and director of the Maxwell Museum, decided to take a break to hunt mountain lions.  (Hibben was an odd guy.)  He chased one particular lion for many miles along the canyons and cliffs, and eventually followed its tracks into a small cliff dwelling high up on a canyon wall.

Looking South from Very Large Array, Plains of San Agustin, New Mexico

When he entered the site, Hibben found to his surprise a pile of bows, which had apparently originally been stacked in a corner but had been largely scattered and broken throughout the room.  Furthermore, there were arrows strewn all across the floor of the site and the cave in which is was situated, also broken.  There were about 94 bows and 4,000 arrows, an astonishingly high number.  While they all appeared to be broken, probably by bears who had used the cave as a lair in the time since its abandonment, the sheer number of specimens made this cache an unparalleled resource for understanding early Southwestern weaponry.  Realizing the importance of his discovery, Hibben collected the bows and brought them back for further study, after which he wrote a short article describing them which was published in American Antiquity in 1938.

The article is short and doesn’t give nearly as much information about the bows and arrows as would be ideal, but it does mention certain characteristics which are important in comparing these bows to other prehistoric and ethnographic examples.  These all appear to have been self bows; Hibben made no mention of any sinew backing or recurving.  They were also large.  The average length of the restorable ones was about four and a half feet, and the longest was almost five feet long.  The shortest was about three feet long, which Hibben described as “too small for any serious use,” although this is a typical length for recurved bows.  Only one bow retained any fragment of bowstring, and it was made of yucca fiber.  Sixteen bows were decorated with red or black stripes.

Railroad Crossing, Very Large Array, Plains of San Agustin, New Mexico

Hibben was unable to determine the type of wood for all of the bows, but from the ones that could be determined it was clear that oak was the preferred material.  Other identified woods were piñon, (ponderosa?) pine, willow, mountain mahogany, and sycamore.  (Note the absence of juniper, which will be important later.)  Interestingly, the bowyers don’t seem to have selected particularly fine staves from which to make the bows, and knots are frequent.  The surfaces,  however, were finely finished and probably polished, suggesting that a considerable amount of effort did go into making these bows.

The most interesting thing about the arrows is that out of the thousands in the cave, only eleven had notches for stone heads to be attached.  The rest were merely sharpened to create wooden points.  This is important to keep in mind, given the large role arrowheads tend to play in theories about prehistoric weaponry.  It’s understandable, since stone projectile points are the most durable parts of any weapon system, but if this ratio of stone to wooden points is typical (and there is of course no way to tell if it is) it suggests that stone points may not actually have been nearly as central as archaeologists often assume.  The arrows were also elaborately decorated in a variety of designs and colors which Hibben interpreted as property marks.

Historic Marker, Plains of San Agustin, New Mexico

There were some sherds of Mimbres Black-on-white pottery associated with the cache, suggesting that it dates to the Classic Mimbres period, ca. AD 1000 to 1150, which would make it contemporaneous with Chaco.  This temporal placement makes sense, since the self bow is known to have been the main weapon used in the Southwest in this period.  The recurved bow was introduced later, perhaps during the period of change and instability in the region around AD 1300, and persisted into historic times.  The persistence of the recurved bow was presumably due to its considerable advantages over the self bow, which faded into the distant past as most examples deteriorated.  One cache, however, survived thanks to its location in a sheltered cave, and was rediscovered due to the efforts of a crafty mountain lion and the archaeologist who pursued it.  Knowledge comes about in odd ways sometimes.

Hibben, F. (1938). A Cache of Wooden Bows from the Mogollon Mountains American Antiquity, 4 (1) DOI: 10.2307/275360

Read Full Post »

Atlatls to Bows: Enter the Bow

Arrowheads at Chaco Visitor Center Museum

I’ve said quite a lot about atlatls, so perhaps it’s time to move on to the second part of this series.  The bow and arrow is a sufficiently popular weapon system even today that it doesn’t need much introduction.  It’s important to note, however, that most archaeologists have concluded that the bow and arrow is quite a complicated system and not at all an obvious idea.  This has led to a general consensus that the bow and arrow was probably invented only once somewhere in the Old World and subsequently diffused gradually across the Bering Strait and southward through the Americas.  This process can be seen pretty clearly in the Southwest, one of the few places where actual examples of both atlatls and bows have been found in well-dated contexts.  Those contexts show that the bow definitely did completely replace the atlatl sometime in the late centuries BC or early centuries AD.  Evidence from Mesoamerica shows that the bow and arrow didn’t arrive there until somewhat later, during the Late Postclassic period, and never really caught on, possibly because of the cultural importance of the atlatl.

Beyond those two well-documented areas, however, we have few datapoints with which to directly track the spread of the bow and arrow and the replacement of the atlatl.  In most (but not all) parts of North America the atlatl was no longer in use at the time of European contact, so complete replacement does seem to have taken place in most areas, but there is little direct evidence for when this happened where.  The use of the atlatl can be seen in some areas through durable artifacts associated with the atlatl-dart complex such as weights, hooks, and finger-loops, but these were not used with atlatls in all areas, and the bow-and-arrow complex has no such associated durable artifacts beyond projectile points.  So while in some, but not all, places we can see roughly when the atlatl stopped being used by noticing the latest occurrences of durable artifacts associated with it, and we can be reasonably sure from ethnohistoric and ethnographic documentation that it was replaced by the bow, we can’t tell if the replacement was sudden or gradual or what mechanism accounted for it.

In some places the atlatl did survive into historic times.  In addition to Mesoamerica and the Andes, where it may have had special cultural status, it was widely used in the Arctic (along with the bow) at the time of the early European expeditions of the nineteenth and early twentieth centuries.  This is interesting because if the bow did indeed diffuse into the Americas from Asia it would have to have come through the Arctic.  This implies that the bow must have been known in that area quite early, and yet it never totally displaced the atlatl.  There is some evidence that the atlatl may have been more useful for fishing and hunting of marine mammals than the bow, which would account for its continued use among the coastal Arctic groups who depended very heavily on those food sources.  I’ll discuss that more in a future post.

Potsherds and a Chipped-Stone Tool, Homol'ovi Ruins

So where did the bow come from, and when was it invented?  Most archaeologists have concluded that the bow and arrow system is sufficiently complex that it is unlikely to have been invented more than once, and thus the idea that it originated quite early somewhere in the Old World and gradually spread is widely accepted.  Not everyone agrees; Oren Evans of the University of Oklahoma published an article in 1957 reporting on some experiments and arguing that, contrary to this general consensus, the basic principle of the bow would have been easy to discover and that it was likely invented multiple times in different places.  He even mentioned having seen a young boy playing around with sticks coming up with the general idea apparently on his own, although he acknowledges that the boy could well have seen bows before, a possibility that rather severely undermines the value of this piece of anecdotal evidence.  The hard part, according to Evans, would have been refining the bow and arrow into an efficient weapon system that would be superior to the atlatl.  This theory has not been widely accepted.

More recently, Malcolm Farmer published an article on the physics of both the atlatl and the bow, arguing on the basis of research by Bob Perkins that the altatl-dart system is actually very similar to the bow and arrow in the physical principles of its operation.  He goes on to argue that they likely originated in the same area, most likely northwestern Africa.  The earliest evidence for atlatl use comes from the distinctive projectile points of the Aterian tradition, dating to around 40,000 years ago (although there is recent evidence that the Aterian may have begun much earlier than previous thought).  Smaller points thought to be arrowheads appear with the later Oranian and Capsian traditions starting around 15,000 years ago.  “Oranian” and “Capsian” are clearly different entities, but it’s not totally clear to me if they are distinct primarily geographically or chronologically.  Also, after seeing the extensive debate over the relationship between point size and weapon type in North American archaeology, some of the pronouncements about these African points strike me as questionable.  This is very much not my area of expertise, though, so I can’t say much more.

I am skeptical about Farmer’s argument that the atlatl and the bow likely originated in the same area because they operate according to similar physical principles.  Even if the implication that the bow and arrow developed out of the atlatl and dart is correct, the atlatl seems to have originated so much earlier than the bow, and spread so far by the time the first evidence for the bow appears, that from a theoretical perspective there’s no reason the change had to occur in the atlatl’s initial home, although that is of course not evidence that it didn’t.  If it is in fact true that the earliest archaeological evidence for both technologies is in northwest Africa, however, and I have no reason to doubt it, then that is much more solid evidence and has interesting implications for cultural continuity and change in that area over a very long period of time (25,000 years or so if the above dates are accurate).  I don’t know of any other theories about the times and places of origin of these weapon systems, and Farmer seems to imply that the matter hasn’t gotten much attention.  I’m not really competent to evaluate Farmer’s theory, but as long as I’m discussing the transition from the atlatl to the bow I figure I should at least mention it as an attempt to understand the origin of both.  I’ll have more on the transition itself in future posts.

Evans, O. (1957). Probable Use of Stone Projectile Points American Antiquity, 23 (1) DOI: 10.2307/277288

Farmer, M. (1994). The Origins of Weapon Systems Current Anthropology, 35 (5) DOI: 10.1086/204331

Read Full Post »

Atlatls to Bows: A Very Strange Atlatl from Washington State

Starbucks, Durango, Colorado

Sometime in the early 1950s a wooden object was dredged from the mouth of the Skagit River, north of Seattle.  It ended up in the possession of Mr. and Mrs. Edwin Johnson, residents of the nearby town of La Conner.  In 1952 the Johnsons showed it to two local archaeologists, Herbert Taylor of Western Washington College of Education (now Western Washington University) in Bellingham and Warren Caldwell of the University of Washington in Seattle.  Taylor and Caldwell immediately recognized it as an atlatl, but it was unlike any atlatl they had ever seen.  It was about 40 cm long, not an atypical length for a North American atlatl, and its width was also typical, but it had on its upper surface a very elaborate carving, 11 cm high, of what looked like a dragon or lizard.  It also had two carved finger holes, similar to some Mesoamerican examples.  They subsequently searched the archaeological literature on atlatls but could find no other examples anywhere of this kind of carving on an atlatl.  Realizing the uniqueness of this artifact made it an important find, they published a short article on it in American Antiquity in 1954, complete with a rather low-quality photograph.  In the article they noted the various possible explanations for the atlatl’s origin, including the possibility that it was a modern fake, which they disregarded because of its uniqueness (if it was a modern imitation, what on earth could it be an imitation of?), and they concluded that its intact condition suggests it was deposited at or near the river’s mouth rather than floating down the river from somewhere inland.  This left two possibilities: either it was a local example of a previously unknown type of artifact, or it was an artifact from elsewhere that somehow ended up in the area in antiquity.  They also noted that atlatls are not known at all from the Pacific Northwest ethnographically.  They didn’t mention the fact that atlatl weights have been found in the Northwest archaeologically, possibly because the main article describing these finds didn’t come until five years later, but since this atlatl doesn’t seem to have any place to attach a weight that wouldn’t be very relevant anyway.  They didn’t present any theories about where the atlatl may have come from or what it meant, saying only that if the Johnsons agreed they would try to do some more detailed analyses to try to get clearer answers.

Starbucks, Philadelphia, Pennsylvania

The British Museum‘s catalog entry for the Santa Barbara atlatl (which has similar finger holes) mentions the radiocarbon dating of a “highly decorated atlatl, dredged up from the Skagit River in the 1950s and dated to the Marpole phase of the Northwest Coast archaeological record,” presumably this one, but the reference is to a 1987 article in the Canadian Journal of Archaeology (with a very uninformative abstract) that I have not been able to access, as Rutgers apparently does not subscribe to the paper version of the journal and the electronic version only goes back to 2004.  I may try to track down a copy of the journal to see the article, since it would be very interesting to see the actual date.  The Marpole phase apparently dates to the late centuries BC and the early centuries AD, although the exact dates given in different sources vary.  In any case, it’s a very mysterious item, and unlike the Santa Barbara example there’s no obviously plausible explanation for it.  It could just be that atlatls like this were common in the Northwest at this time but since they were made entirely of wood none of the others have survived in the very wet climate (probably the most likely explanation).  The idea that it was an import from somewhere else is tempting given its uniqueness, but where could it have come from?  Interestingly, the only atlatls I know of with elaborate carving like this on the top are some ethnographic examples from New Guinea in the Smithsonian, but that’s got to be a coincidence.  Anyway, this is just one more of the many odd little stories in the history of atlatl research.

Taylor, H., & Caldwell, W. (1954). Carved Atlatl from Northwest Coast American Antiquity, 19 (3) DOI: 10.2307/277136

Read Full Post »

Atlatls to Bows: A Very Strange Atlatl from California

Higuera Street, San Luis Obispo, California

In November of 1793 a British naval expedition commanded by Captain George Vancouver arrived at the small Spanish settlement of Santa Barbara on the coast of California.  Vancouver’s primary mission was to explore and map the poorly understood northwest coast of North America, building on the more preliminary information provided earlier by Captain James Cook.  He was quite successful at this, and the detailed maps produced by his expedition greatly enhanced British understanding of this area, which was becoming very important geopolitically as a result of its potential resources and increasing competition among Britain, Spain, and Russia to claim it.  When he arrived at Santa Barbara he was headed south, having spent the spring and summer exploring the area around the island that now bears his name and bound for Hawaii to spend the winter.  He anchored off of Santa Barbara for eight days to rest and resupply, and his men took advantage of the opportunity to trade with the local Spanish and Chumash inhabitants.  Mission Santa Barbara was only a few years old, having been established in 1786, and the presidio where Spanish soldiers were garrisoned was only four years older than that.  Although the Chumash had been in contact with the Spanish since the Cabrillo expedition of 1542, the permanent Spanish presence in their territory dated only to the establishment of Mission San Luis Obispo in 1772, and at the time Vancouver’s expedition stopped by they were only just beginning to move to the missions and experience the profound and complicated cultural changes that would result.

Chamber of Commerce, San Luis Obispo, California

George Goodman Hewett, Surgeon’s First Mate on Vancouver’s flagship, HMS Discovery, was among the members of the expedition who did some trading with the locals at Santa Barbara.  Hewett apparently had a strong interest in the customs and lifestyles of the various peoples the expedition encountered, and he collected from them various items of material culture whenever possible.  Over the course of the four years that the expedition ended up taking he acquired a substantial collection.  While the greatest number of items in the collection were from the places the expedition spent the most time, particularly Hawaii and the Pacific Northwest, the items from Santa Barbara were (and are) of particular interest to later anthropologists, since Hewett collected them at a time when traditional Chumash culture, now known primarily from the very detailed but nevertheless retrospective ethnographic fieldwork of John Peabody Harrington in the early twentieth century, was still mostly intact and only beginning to be affected by missionization and Spanish contact.  Hewett’s collection remained in his family until 1891, when it was acquired by the British Museum, where it remains.  A description of some of the most significant items was published by Charles H. Read in 1892.

Hill, San Luis Obispo, California

Read’s description included two atlatls.  One is an elaborately carved example from southeast Alaska, where use of the atlatl is known to have persisted into modern times, perhaps because of its usefulness in fishing and maritime hunting compared to the bow.  The other is from Santa Barbara.

First Bank of San Luis Obispo, San Luis Obispo, California

The Santa Barbara atlatl, as it has been known ever since Read’s publication, is very odd for a number of reasons.  For one thing, just at the outset, the idea of the atlatl being used at Santa Barbara in the late eighteenth century is odd.  California is one area where the replacement of the atlatl by the bow and arrow, whenever it happened, is widely agreed to have been complete by long before then.  While atlatl hooks have been found in early archaeological contexts in various parts of the state, including the Chumash area, there is no evidence of atlatl use from later prehistoric contexts, and the copious ethnohistoric and ethnographic literature on the Chumash nowhere mentions the atlatl, whereas the importance of the bow and arrow is discussed many times.  Read was not in a position to know any of this, of course, since this understanding of the culture history of the area came long after his time.

San Luis Surf, San Luis Obispo, California

Furthermore, the form of the atlatl itself is odd.  While archaeological and ethnographic examples from North America, including the Alaskan ones from the Hewett collection, are typically long and thin, the Santa Barbara specimen is short and thick.  While most atlatls are about half a meter long, this one is only 15 cm long, and nearly as wide.  Furthermore, while North American examples typically have either attached leather finger loops or none at all, this one has two large holes carved out of the wood itself.  The wood seems to be a local type, and no one has questioned the authenticity of the specimen or Hewett’s account of its origin (Read notes that Hewett’s record-keeping was pretty good by the standards of his time), but it’s all very odd and hard to explain.  If it represents a survival from a local atlatl tradition, this tradition is suspiciously absent entirely from both the archaeological and ethnographic records.  While it’s true that atlatls, being made of wood, rarely survive archaeologically, this one does have a bone hook, so if it represents a survival of an ancient atlatl type that continued in use after the adoption of the bow and arrow it would be reasonable to expect at least one similar bone hook to survive somewhere, and this still doesn’t address the lack of ethnographic evidence.

Street Signs, San Luis Obispo, California

Nevertheless, the Santa Barbara was generally accepted as an unusual but indigenous type of atlatl until 1938, when the prominent California archaeologist Robert Heizer published an article looking at the issue and coming to a quite different conclusion.  Heizer pointed out the lack of any other evidence for this type of atlatl as well as all the oddities of the Santa Barbara specimen compared to other examples, and went a step further by noting that it bore a striking similarity to the atlatls still in active use at that time by the Tarascans of western Mexico (remember them?).  These also have paired finger-holes carved out of the wood, and have the same widening of the body of the atlatl around the holes.  The dimensions are still different; the Tarascan examples are much longer and thinner than the Santa Barbara one.  There is still a remarkable similarity, however, and Heizer goes on to point out that the Spanish were known to use Tarascans and other Indians from previously colonized parts of Mexico as settlers on the frontier, particularly in the northwest, which is where the expeditions that colonized California in the 1770s are known to have started.  Although there is no direct evidence that the Spanish soldiers and missionaries in California were accompanied by Mexican Indians, given typical Spanish practices it would not be a surprise.  This, combined with the striking similarities between the Santa Barbara atlatl and Tarascan ones, leads Heizer to propose that the Santa Barbara example is not a survival at all, but a reintroduction of the atlatl to the area from Mesoamerica, where it remained in use long after the Spanish conquest.  The Santa Barbara one is clearly of local manufacture, however, which suggests that this process did not simply involve Tarascans bringing their own atlatls to California, although that was presumably part of it.  Rather, once the Mexican Indians were there, they apparently showed the Chumash the use of the atlatl, which they used for fishing and hunting in maritime settings, and the Chumash (who were a coastal people very oriented toward the sea) were sufficiently impressed to copy it themselves.  Since it apparently did not become established securely enough to be noticed or mentioned by either the Spanish or later ethnographers, the Chumash don’t seem to have ultimately decided to adopt it as a core part of their culture or subsistence system, but they do seem to have at least tried it out.  Indeed, Hewett may have encountered the Chumash at a time of experimentation connected to the changes associated with the transition to mission life, and his acquisition of the atlatl may have preserved a moment in time, a tentative embrace of foreign technology that was ultimately rejected and that would therefore otherwise be unknown to history.  Along the same lines, it’s worth wondering why the Chumash were willing to part with this obviously unusual and presumably rare item when all the other things they gave Hewett were rather typical and plentiful items such as bows.  Was whoever tried to copy the Mexican atlatls, or whoever had tried to use the copy made by someone else, displeased with how the experiment had turned out and eager to get rid of the item when a foreigner interested in buying random things showed up?  There’s no real way to tell, of course, and it’s also possible that atlatls like this were used successfully for a while around this time then abandoned for some other reason.  This item is, however, an intriguing window into a complicated past, and it shows that it’s important to look carefully at the stories behind artifacts before constructing theories based on their characteristics.

Heizer, R. (1938). An Inquiry into the Status of the Santa Barbara Spear-Thrower American Antiquity, 4 (2) DOI: 10.2307/275985

Read, C. (1892). An Account of a Collection of Ethnographical Specimens Formed During Vancouver’s Voyage in the Pacific Ocean, 1790-1795 The Journal of the Anthropological Institute of Great Britain and Ireland, 21 DOI: 10.2307/2842277

Read Full Post »

Atlatls to Bows: Loopy

Atlatl Petroglyph Showing Finger Loops, Atlatl Rock, Valley of Fire State Park

Intact atlatls are rarely found, but when they are it’s usually in the Southwest or the Great Basin, arid regions with good preservation conditions for perishable materials like wood and leather.  Some, but not all, of the examples that have been found in these areas have pieces of leather attached as apparent finger loops to secure the atlatl during use.  Since these would be highly perishable even under relatively good conditions for preservation, some have suggested that many of the surviving atlatls that don’t have them now would have originally had them, but this is not universally accepted and there is no real way to tell for sure.  In many other regions there is no evidence of atlatls ever having had finger loops, and they certainly wouldn’t have been necessary in any absolute sense for effective use of the device.

Entrance Sign, Valley of Fire State Park, Nevada

There is evidence that the use of finger loops was somewhat more widespread, however.  Just how widespread they were was discovered by Gordon Ekholm of the American Museum of Natural History, and the interesting story of how he figured it out is documented in an article he wrote that was published in American Antiquity in 1962.  Ekholm had done some excavation in northeastern Mexico and found there some enigmatic artifacts, generally made of shell and crescent- or U-shaped, with drilled holes at the ends.  These had been found by archaeologists in other parts of Mexico as well, especially in the west and northwest, and while no one was quite sure what they were most researchers had considered them ornaments, perhaps used for personal adornment on the nose or, since they often occurred in pairs, on the ears.

Visitor Center, Valley of Fire State Park

Then, in the summer of 1960, Ekholm was visiting the British Museum and saw on display there a very elaborate atlatl from Mexico with two crescent-shaped shell finger loops elaborately tied to it with cotton thread.  When he saw it he immediately realized that the mysterious shell crescents must have been finger loops for similar atlatls.  He got in touch with the museum’s curator of ethnography, who examined the atlatl more closely and gave Ekholm more information on it.  It turned out that only one of the loops was original, the other being a replacement, but the binding for the original one was apparently original as well.  Inspired by this discovery, Ekholm sought out information on other atlatls in European museums that had such finger loops.  He found only one, at the Pigorini Museum in Rome.  The director of the museum provided him with photographs of the atlatl and confirmed that, contrary to published information, the loops were made of shell rather than bone.  Both loops on this example were apparently original, as was the binding.

Bighorn Sheep Petroglyph Panel, Valley of Fire State Park

Once he had confirmed that the examples in London and Rome were similar to the artifacts he knew from Mexico, Ekholm looked through the collections of the American Museum and the Heye Foundation’s Museum of the American Indian (now the National Museum of the American Indian), as well as in the published literature, for examples of the artifacts to test his hypothesis.  He found that they generally fit it very well, although a few seemed a bit small to be finger loops.  These, however, could be explained by the fact that the bindings on the European specimens were rather loose, leaving a bit more space between the inside of the loop and the wood of the atlatl than would be apparent from just looking at the loop itself.  The only example he found that was notched in a way suggesting it was attached directly to the atlatl was an unusual specimen from Sonora made of green stone, and he suggested that the atlatl itself may have been carved to narrow it a bit where the loops were attached.  Geographically, the examples Ekholm could find were mainly from the western part of Mexico, although there were also some from the northeast and from central Mexico, including some examples from Teotihuacan.  The only examples from the Maya area were unusual in size or shape and may have been some other sort of artifact that only superficially resembled these finger loops.  In general, this type of atlatl seems to have been primarily associated with the north and west of Mexico.  A subsequent paper by Ann Johnson of the University of Kansas added two more examples from Sonora which resembled Ekholm’s unusual notched example and suggested that this might indicate a specific regional tradition on the far northern fringe of Mesoamerica.  She notes in this connection the leather finger loops on Southwestern examples, which tend to have a narrowing in the atlatl itself where the loops were attached, just as Ekholm had suggested for his Sonoran loop.  Johnson also suggested that another enigmatic artifact type, the small so-called “cruciform” artifact made of stone and shaped like a cross or a four-pointed star, might be an atlatl attachment, since these were typically found in rows associated with male burials.  To support this idea she relied mainly on evidence from the Aztec codices suggesting that some atlatls were decorated with star symbols, which seems rather weak evidence to me, but the idea is at least plausible.

Petroglyph Sign, Valley of Fire State Park

Ekholm noted at the end of his paper that the importance of these finger loops extended far beyond identifying a mysterious artifact type.  Since most Mesoamerican atlatls were apparently made entirely of perishable materials, and since most of Mesoamerica is not conducive to preservation of such materials, finding evidence of atlatl use in Mesoamerica is very difficult and tends to rely mostly on artistic and textual sources, which can be difficult to interpret.  With the discovery of these durable loops, however, Mesoamerican archaeologists gained a marker of atlatl use similar to the weights and hooks found in various parts of North America.  Clearly these loops were only used on a particular type of atlatl, probably one of many in use in Mesoamerica, but having a reliable marker for one type is better than having nothing, which is what Mesoamerican archaeologists had before.  The possible connections between the leather finger loops of the Southwest and these shell and stone loops, most common in the area in between the Southwest and the Mesoamerican heartland, would be interesting to look at, although I’m sure it’s a complicated issue and there may not be any straightforward answers.  This little chapter in the story of atlatl research is interesting in and of itself, however, and it shows the importance of serendipity in making connections between things that have been lying in plain sight without their importance being recognized.

Ekholm, G. (1962). U-Shaped “Ornaments” Identified as Finger-Loops from Atlatls American Antiquity, 28 (2) DOI: 10.2307/278375

Johnson, A. (1971). Finger-Loops and Cruciform Objects American Antiquity, 36 (2) DOI: 10.2307/278671

Read Full Post »

Atlatls to Bows: The Hook Brings You Back

Post Office, Independence, California

Atlatl weights are the most widespread attachments to atlatls that are durable enough to survive in conditions where the wooden parts decay, but they’re not the only attachments known to have been used.  Another type of attachment, of more obvious function though of much more limited range, is the “hook” or “spur” near the back end of the atlatl that cradles the nock of the dart when it is being thrown.  In Mesoamerican and Southwestern examples, the hook is generally gouged out of the wood of the atlatl itself, either flush with the upper surface (with a recessed groove for the dart) or protruding above it with the dart resting at the level of the atlatl surface.  This was probably also the case for most atlatls in areas where they have not survived but their presence is attested by weights.  In a few places, however, hooks were made of stone, bone, or antler, and have thus survived where the atlatls they were attached to have not.

California Welcome Sign

The two main areas where durable hooks were used are widely separated geographically, though not necessarily temporally.  These are California, especially the Central Valley, and the Ohio River watershed, especially the areas south of the river in Kentucky and Tennessee.  In both areas the hooks seem to date to the Archaic period, but beyond that there are few similarities.  The California examples are described in a 1969 article by Francis Riddell and Donald McGeein which classifies them into three types.  The vast majority fall into their Type II, which are shaped like elongated acorns and made usually of bone, although a few examples are of stone.  They come mainly from the Central Valley, especially the Sacramento area, although a fair number come from the San Francisco Bay area and the Santa Barbara Channel as well and a few are reported from other areas such as Los Angeles County and Morro Bay in San Luis Obispo County.  The examples from documented contexts all seem to be associated with the Middle Horizon period, but many are from undocumented excavations or private collections, and their original contexts are unknown.

Morro Bay, California

Riddell and McGeein also define two other types.  Type I appears to be earlier than Type II and consists of hooks shaped like snake heads and made of stone, perhaps exclusively, which makes them quite different from Type II hooks, which are acorn-shaped and usually made of bone.  Type I examples resemble hooks found in Nevada and are found in the Sierra Nevada area on the edge of the Great Basin, as well as in the Central Valley.  Type III is represented by a single specimen from a private collection which apparently contains other similar examples, all from the San Joaquin Valley.  It closely resembles Type I but is considerably larger and made of bone rather than stone, and Riddell and McGeein note that it may be a variant of Type I rather than a separate type.  They also suggest the possibility that Type III is intermediate between Types I and II, assuming the main difference between them is temporal.

Sierra Nevada, Independence, California

Interestingly, Riddell and McGeein note that Middle Horizon sites tend to lack atlatl weights (which in California are often called “boatstones”), although they commonly have Type II atlatl hooks.  This suggests that Type II hooks may have been used on a particular type of unweighted atlatl.  Since there appear to be no intact examples of this type of atlatl, however, due to the poor preservation conditions in the humid environments where Type II hooks are found, they can only offer this as a tentative suggestion.  Type I examples, however, are often associated with weights, suggesting that they come from a different (earlier?) type of atlatl that was weighted.

Mt. Whitney Administrative Office, Lone Pine, California

On the other side of the country are the rather different atlatl hooks known from well-preserved atlatls found in rockshelters in Kentucky and Tennessee.  These hooks are typically made of antler or bone, and the antler ones are often made from an antler tip only slightly modified to be attached to the atlatl and to hold the dart.  One example from Ohio is illustrated in a short article in American Antiquity, but the JSTOR scan of the page is unfortunately of very poor quality and it’s not possible to see any details in the picture.  Although these appear to be of Archaic date, making them roughly contemporaneous with the California examples, they are sufficiently different in form that there is unlikely to be any connection, and independent innovation in both areas is more likely than any sort of contact or diffusion.  The fact that the idea of a durable atlatl hook seems to have been unknown in the Southwest or on the Great Plains also suggests independent innovation.  The practical value of a strong support for the nock of the dart seems pretty clear, and may explain why at least two separate groups came up with the idea, but the fact that it didn’t spread much further in any case and was not invented independently in more cases suggests that it may not have been a major improvement over the more typical gouged hook.

Goslin, R. (1944). A Bone Atlatl Hook from Ohio American Antiquity, 10 (2) DOI: 10.2307/275117

Riddell, F., & McGeein, D. (1969). Atlatl Spurs from California American Antiquity, 34 (4) DOI: 10.2307/277746

Read Full Post »

Atlatls to Bows: Those Puzzling Weights

US and Arizona Flags, Kayenta, Arizona

Most of what we know about prehistoric North American atlatls comes from the many well-preserved examples found by Alfred Kidder and Samuel Guernsey in the early twentieth century in Basketmaker II rockshelters near Kayenta, Arizona.  We know much more about atlatl use in Mesoamerica, where the atlatl was still widely used in the contact era, from descriptions by the early Spanish conquistadors and missionaries, and one account of the De Soto expedition indicates that the atlatl was still used as a weapon in the lower Mississippi valley in the sixteenth century as well.  Elsewhere in North America, however, the bow and arrow had apparently completely replaced the atlatl long before European contact (which in many areas was quite late), so only archaeological evidence shows that it was ever used at all.  In most regions conditions were not good for preservation of perishable materials like wood, so atlatls themselves very rarely survive except in the dry Southwest.  Many Southwestern atlatls, however, had attachments made of more durable materials, and similar artifacts found in other regions indicate that atlatl use was quite widespread throughout the continent.

US Highway 160, Kayenta, Arizona

The best known and most mysterious of these attachments are the so-called “atlatl weights” found securely tied or strapped to many Southwestern atlatls.  These are typically small pieces of stone, rounded or rectangular in shape, secured to the underside of the atlatl between the finger loops and the groove where the dart was placed.  Once they were identified as atlatl attachments from the intact Basketmaker specimens, similar artifacts in many other areas (such as the northern Great Plains and the Pacific Northwest) were grouped with them as well.  Since they are made of stone rather than wood, they preserve much better than the atlatls they were presumably attached to, so assuming the identification of their function is correct they constitute a reliable indicator of the distribution of weighted atlatls throughout North America.

Parking Lot at Hampton Inn, Kayenta, Arizona

The question of what these things actually were remains, however.  They are conventionally described as “weights,” and it’s certainly true that they weigh down the atlatl, but beyond that it is not at all clear why they were attached.  It’s important to note in this context that weights are not found on atlatls elsewhere; Mesoamerican examples, in particular, never have them, and are typically made of solid pieces of wood.  Since the atlatl was very important in Mesoamerica, it is interesting that the Mesoamericans never seemed to think it worthwhile to add weights.  This suggests that the weights on North American examples are either connected to some functional or design difference between types of atlatl or non-utilitarian.  Some of the attachments on some of the Basketmaker examples seem likely to have been charms, but others are more substantial and look more like they were added for a functional reason.  The exact nature of that reason remains elusive, however, and not for lack of attention from researchers.

Power Lines and Trailers, Kayenta, Arizona

One early experimental study aimed at determining the function of the weights was done by Orville H. Peets and published in 1960.  Using a replica of a well-preserved weighted atlatl found in a rockshelter in West Texas, Peets found that the weight made no significant difference to either the distance the dart traveled or the force with which it did so.  This is in contrast to the assumption by many earlier researchers that the additional weight improved force and/or distance.  Peets concluded that the purpose of the weights was instead to keep the atlatl and dart in balance on the hand while throwing, and he noted that Mesoamerican atlatls, which do not have weights, are typically heavier and would stay in balance more easily than the thin Southwestern examples.  Calvin Howard’s later experiments, which I mentioned earlier, didn’t focus primarily on the issue of weights, but they did support Peets’s conclusion that the weights did not improve the distance or force of the throw.  Indeed, Howard found that adding the weight decreased the distance substantially.

"No Water Hauling" Sign, Kayenta, Arizona

A later article by John Palter took issue with Peets’s conclusion that the weights served to balance the atlatl by pointing out that in Australia, where the atlatl was widely used well into the era of European contact, very large darts were thrown easily with no attempt to bring the atlatl/dart system into balance on the hand, which suggests that balance actually doesn’t matter.  Palter further argued that the fact that Southwestern atlatls, unlike Mesoamerican ones, were typically thin and flexible may indicate that flexibility was an important consideration in their design, and that the weights may have served to amplify this flexibility.  He doesn’t give much support for this other than pointing out the use of flexible (but unweighted) atlatls from Australia, and he does acknowledge that at least some of the items attached to Basketmaker atlatls were likely charms with no practical function.

Horse Trailers, Kayenta, Arizona

Yet another article on the issue of atlatl weights was published by Anan Raymond in 1986.  Using high-speed photography to carefully document the actual motions involved in throwing the atlatl, Raymond concluded that Howard’s description of how it works was correct and that the spur of the atlatl, where it contacts the nock of the dart, essentially moves in a straight line forward during the throw.  He also notes that much of the inconsistency in previous experimenters’ conclusions about the function of weights is likely due to differences in the design of the atlatls they used, most of which were made of modern materials and not necessarily comparable to archaeological examples.  (He mentions Peets as an exception.)  His own experiments were done with replicas of Basketmaker atlatls and darts using materials as close as possible to the originals.  He found that the weights did improve both speed and distance, but by a very small amount that may not have been statistically significant.  He did notice in the high-speed photography that the atlatl did flex during the throw, and he speculated that, as Palter predicted, the addition of a weight might enhance the flexing, but the film speed was too slow to give enough frames of the flexing to see any difference between weighted and unweighted atlatls.

Kayenta Presbyterian Church, Kayenta, Arizona

Since his experiments showed that any improvement in speed or distance from the addition of a weight were likely negligible in practice, Raymond took a different tack in trying to explain the function of the weights.  He argued that the addition of weights increased the inertia of the throwing system, which made throwing more difficult but also increased angular momentum, which kept the throwing arc more stable and thus improved the accuracy of the throw.  This is an interesting and plausible explanation, and I think it makes more sense than any of the other theories offered by earlier researchers.  After all, in either hunting or war accuracy is a very important consideration, and a substantial increase in accuracy could easily be worth a small decrease in distance or force.  (Raymond also did some interesting comparative experiments to look at the differences between atlatls and bows, which I’ll discuss further in another post.)

Water Tower, Kayenta, Arizona

Even if Raymond was right that it was accuracy rather than force or distance that made weighted atlatls advantageous, there are still some puzzles remaining about the weights.  Most importantly, if they were in fact so useful, why didn’t they spread further?  They do seem to have been used over most of North America, but in Mesoamerica and further south people seem to have achieved more or less the same thing by just making heavier (and less flexible) atlatls.  The differences in types of atlatls used in different geographic areas, and the persistence of atlatl use even after the introduction of the bow and arrow to some, but not all, areas, suggests that a wide variety of cultural factors were involved in choices about weaponry.  Given that, there’s a limit to how much information can be obtained from looking at these issues from a purely technological perspective, as valuable as that perspective undoubtedly is.

Bushnell, D. I. Jr. (1905). Two Ancient Mexican Atlatls American Anthropologist, 7 (2), 218-221 DOI: 10.1525/aa.1905.7.2.02a00040

Butler, B., & Osborne, D. (1959). Archaeological Evidence for the Use of Atlatl Weights in the Northwest American Antiquity, 25 (2) DOI: 10.2307/277441

Fenenga, F., & Wheat, J. (1940). An Atlatl from the Baylor Rock Shelter, Culberson County, Texas American Antiquity, 5 (3) DOI: 10.2307/275282

Howard, C. (1974). The Atlatl: Function and Performance American Antiquity, 39 (1) DOI: 10.2307/279223

Neuman, R. (1967). Atlatl Weights from Certain Sites on the Northern and Central Great Plains American Antiquity, 32 (1) DOI: 10.2307/278777

Palter, J. (1976). A New Approach to the Significance of the “Weighted” Spear Thrower American Antiquity, 41 (4) DOI: 10.2307/279019

Peets, O. (1960). Experiments in the Use of Atlatl Weights American Antiquity, 26 (1) DOI: 10.2307/277169

Swanton, J. (1938). Historic Use of the Spear-Thrower in Southeastern North America American Antiquity, 3 (4) DOI: 10.2307/27562

Read Full Post »