#lang scribble/base
@(require "../Bibliography/cite.rkt"
          scriblib/footnote
          "../Images/image-support.rkt")

@define-cite-footnote[footnote
                      make-footnote
                      cite-footnote]

@title{Introduction}

@author{Morgan Lemmer-Webber}

@;section{Women and Textiles in Roman Myth}

@;The association between women and textile production has a long and rich history in the western world. Imagery of women spinning within their homes was pervasive in Greek literature and art and reflected the lives of women in that culture. In spite of the common perception that Roman textile production took on increasingly commercial properties, this association persisted. Mythological and heroic stories from Greek literary traditions in particular were revisited by early Imperial authors. While textile production was associated with several Goddesses, the most prominent were the Parcae and Arachne. Their attributes--a spindle and a loom respectively--reflect the two aspects of textile production that are most commonly associated with women: spinning and weaving. 



@section{Central research questions}

The moral ideal of a productive and virtuous wife is embedded in Greek and Roman mythology and widely recorded in ancient literature and visual traditions. Archaeological evidence from domestic contexts confirms that textile production was to an extent carried out in the home, at times alongside larger commercial practices. Epigraphic and literary evidence suggest that the association between women and wool working remained prevalent throughout Roman culture. In the pre-Roman through early Republican periods, textile production was a domestic activity performed by women and slaves. It is generally accepted by scholars that the textile industry shifted into an increasingly commercial endeavor in the Republican and Imperial periods and therefore women's role in textile production became less prominent. However, the performative and idealized associations between women and textile production persisted in Roman culture.

@; Do I have to push the parameters past the third century? Verify with case study chapters

In this study I will examine the persistence of this tradition and explore the roles women played in wool working in the Roman Empire from the first through third centuries CE. The evidence for Roman textile production in general is relatively scarce and scattered, especially when it comes to women's involvement. There is no one source or site that can give us a comprehensive view of women's roles in textile production. Instead, we see several emerging themes including the ancient idea of the productive matron; the trope of the dutiful woman; the labor force within the Roman household; and the labor force in a @as-index{cottage industry}. Written and epigraphic sources that discuss the topic either tend to be skewed toward male involvement in the larger textile economy or toward the domestic roles of women. Addressing the above themes as well as the missing evidence, I argue that women played integral roles throughout the Roman textile industry.

In order to narrow the focus of this research, I will use the sites of Karanis, Trier, and Ephesus as case studies. I chose these sites because they each provide multiple sources and types of evidence that have been published. These sites also reflect a geographic dispersal throughout the Empire. 'Roman culture' was not a consistent construct. There is wide diversity throughout the empire due to the local cultural traditions and by assessing sites from different provinces, I hope to gain some insight into both that diversity and the consistencies. As in any study that cross-analyzes disparate sites across the empire, I will also explore what impact, if any, the long-distance trade network that operated throughout the empire had on textile production and the role of women in the provinces. Therefore, while the scope of this research focuses on evidence from the first through third centuries CE, a brief summary of pre-Roman evidence of textile production will be provided for each site.

@; Is this still applicable?: John Peter Wild has argued that Roman textile production largely retained its iron-age characteristics in the western provinces, was influenced by the technological advances from Persia and Greece in the eastern provinces, and reflected a mixture of the traditions within Italy.@cite-footnote[Wild2002] By focusing on evidence from three different provinces within the empire, I hope to test this hypothesis.

This study will begin with three brief case-study chapters that present archaeological, epigraphic, and visual evidence or lack thereof for textile production at each case study site. These will be followed by three synthetic chapters that place the evidence from these sites into the larger narrative of Roman textile production. In chapter one I will discuss the wealth of material culture of textile production supported by epigraphic evidence in the form of apprenticeship contracts from Karanis. In chapter two I will cover evidence from Trier with a focus on visual representations of the textile industry through the Igel column as well as references to professional textile associations. In chapter three I will discuss the archaeological evidence of textile production in both domestic and funerary contexts from Ephesus, particularly the ornate @index['("distaff")]{distaffs} that suggest a symbolic or ceremonial use. Chapter four will discuss how evidence from these sites reflects the domestic production of textiles as evidenced more broadly within the Roman empire. Chapter five will consider the commercial production of textiles while also calling into question how distinct the division of domestic and commercial production of textiles actually was in the Roman world. Chapter six will look at the performative aspects of women and textile production particularly in the contexts of marriage and funerary rites.@footnote{In this context, I am using the term performative to refer to rituals or traditions in which textile tools and production are used as a symbol or attribute referencing traditional feminine virtues and roles.} This evidence will be analyzed with an emphasis on how changing political, economic, and social factors impacted changes in practice. This interdisciplinary approach allows for a more nuanced view of how textile manufacture related to women's lives. The conclusion will then situate the evidence from these three sites into a discussion of women's roles in textile production and whether evidence from the provinces reflects the cultural ideals of femininity of the Roman Empire more broadly.

@; NDC feedback: Is there a diachronic or evolutionary aspect to it? Do you want to explore the impact of commercial professional production on women's lives?

Through this structuring of chapters, I hope to demonstrate how the association between women and textile production manifested itself over time and in various cultural climates. In fields where evidence is scarce and varied it is necessary to analyze where and why certain streams of evidence were preserved and how that informs our interpretations. For this reason, highlighting the lack of evidence for specific themes in different locations may provide insight otherwise ignored in the scholarly discourse of textile production. 

@section{Stages of Roman Textile Production}

Overall, this study focuses on the social and cultural aspects of textile production rather than the technical processes themselves; however, a broad overview of the stages of production is in order. As in every type of production, the first obstacle is obtaining raw materials. It likely varied by location whether the sheep were raised in town, in the surrounding hinterland, or whether the wool was imported. In addition to faunal evidence of the sheep themselves, the archaeological record may include the metal sheers used to remove the excess winter @as-index{fleece} annually in the spring.@cite-footnote[Wild2002 29] The @as-index{fleece} is then cleaned and processed into long bundles of fiber called a roving.@footnote{This stage of processing may have included combing the wool so that the fibers faced in a uniform direction, though this step was not necessary, @cite[Wild1976 168]}

While it is possible to spin directly from the roving, the Romans often wrapped the roving around a @index['("distaff")]{distaff} to help manage the unspun quantity of wool.@cite-footnote[Lipkin2012 37] The primary element to the distaff is a shaft to wind the roving around, this might have a fork at the end to help secure the wool. The style of distaff that most commonly survives in the archaeological record from Roman contexts is a ring-distaff or finger distaff, which involves a shaft with one end terminating in a ring and the other end often topped with a decorative element. The shaft may be decorated with simple patterns, the texture of which help adhere the roving to the distaff without it slipping. When the distaff is loaded with wool, the decorative element at the top would be visible above the wool. Since most of the surviving distaffs are made of glass, bone, ivory, jet, or other precious or semi-precious stones, their dominance may represent a survival bias over simpler distaffs made of wood.@cite-footnote[Wild2002 9] The hard nature of these materials combined with the cushion of the roving also are unlikely to show the same signs of wear that a wooden distaff would.@footnote{It is possible that the movement of the wool over stone and bone @index['("distaff")]{distaffs} would have created a higher polish or sheen, but this is hard to determine.} Combined with the fact that they are frequently found in funerary contexts, it is difficult to determine if these surviving examples were ever used or were purpose-made to serve as grave goods.

The basic tool required for spinning the processed wool into thread is a drop @index['("spindle/whorl")]{spindle}. This consists of a shaft and whorl. Whorls come in multiple shapes and sizes depending on the desired size, material, and consistency of the thread. The basic description of a whorl is a round or disc-shaped object that is heavier than the shaft with a hole in the center. The weight of this @index['("spindle/whorl")]{whorl} gives the spindle the centrifugal force to continue spinning while the @index['("quasillaria/spinner")]{spinner} drafts out the thread. The shaft of the spindle may be slightly tapered in order to secure the whorl to it. Though not strictly necessary to an experienced @index['("quasillaria/spinner")]{spinner}, a groove, notch, or metal hook may be included at the top of the spindle to secure the thread and ensure that it does not unravel. Spindle shafts were typically made of wood and therefore do not often survive. Spindle whorls could be made of wood, clay, glass, stone, or ivory and therefore are more common in the archaeological record. Very few metal hooks that were potentially from @index['("spindle/whorl")]{spindles} survive, suggesting that they were not the norm. Where wooden spindle shafts survive, signs of wear include ridges or incisions near the top of the spindle from the thread wearing against the shaft, marks around where the whorl would have sat, and general nicks and scrapes. On @index['("spindle/whorl")]{spindle} whorls, the most common signs of wear are chips around the edges of the whorl.

@;Add photo of my own spindle with signs of use

To make thread, the @index['("quasillaria/spinner")]{spinner} holds the @index['("distaff")]{distaff} in one hand and uses the other hand to set the @index['("spindle/whorl")]{spindle} twirling. She then uses that free hand to draft the thread out by pulling small clusters of fiber out of the roving at a time. With her thumb and forefinger she controls the amount of twist that travels up to create the thread.@cite-footnote[Wild1976 170] Since both hands are occupied, she may use her teeth to remove impurities or inconsistencies from the thread.@footnote{ According to Catullus, @italic{Carmina} 64 305-320. However, it is unclear whether this would have been a common practice among @index['("quasillaria/spinner")]{spinners} or if Catullus was using it as a literary device. The impurities in wool would have consisted of whatever the sheep got into such as straw, dirt, and feces. Catullus is not known to shy away from visceral imagery.} Once the @index['("spindle/whorl")]{spindle} loses momentum and stops spinning, she grabs the suspended spindle, winds the thread around the spindle shaft, and repeats the process. In his description of the Parcae, the Roman equivalent of the Greek Moirai who determine each person's fate by spinning, measuring, and cutting the thread of their life,@cite-footnote[Guilleux2016 8] Catullus gives a detailed description of this process:

@nested[#:style 'inset]{... their hands pursued their never-ending toil, as of custom. The left hand bore the @index['("distaff")]{distaff} enwrapped in soft wool, the right hand lightly withdrawing the threads with upturned fingers shaped them, then twisting them with the prone thumb it turned the balanced @index['("spindle/whorl")]{spindle} with well-polished whirl. And then with a pluck of their tooth the work was always made even, and the bitten wool-shreds adhered to their dried lips, which shreds at first had stood out from the fine thread. And in front of their feet wicker baskets of osier twigs took charge of the soft white woolly @as-index{fleece}.@footnote{Catullus, @italic{Carmina} 64 305-320}}

@; CW: Also wonder if the footnote's citations can be updated, but probably not urgent.

@;The Parcae, the Roman equivalent of the Greek Moirai, determine each person's fate by spinning, measuring, and cutting the thread of their life.@cite-footnote[Guilleux2016 8] Nona, as the Parca who spun the thread, is identified by her spindle. In his description of the Parcae in @italic{carmen} 64, Catullus gives a detailed description of the act of spinning.@footnote{See also: Ov. @italic{TR.} 1.63, Hor. @italic{Od.} 2.3.16} Using a distaff in their left hands to hold and control the wool roving, they spin the thread of life using a well-polished spindle. Their right hands alternate between turning the spindle and drafting, pulling out a controlled number of fibers from the roving to spin into an even distribution. Since both hands are occupied, they use their teeth to remove impurities or inconsistencies from the thread. The wool baskets at their feet held more raw wool fleece, ready to be combed and processed for future spinning projects. In spite of this detailed procedural description of textile production, the primary role of the Parcae in the poem is to sing the fates of the couple at the marriage of Peleus and Thetis. Following the above passage they commence with their song which is interspersed frequently with the refrain "Run, drawing the thread, run spindles!"



The combination of the velocity of the spin, the weight of the whorl, and the number of fibers she drafts at a time determine the thickness of the thread. Most surviving examples of Roman textiles use only single threads created by the above process. Some examples, however, use a sturdier plied thread either for the warp strands or sewing thread.@cite-footnote[Wild2002 10] To ply thread, two or more single threads are spun in the opposite direction (i.e. if they were spun clockwise for the initial single, they would be plied counter-clockwise), causing the threads to fold back on each other. Spinning is also the most time-intensive step of the textile process and therefore can become a production bottleneck. While we have no primary records indicating the actual amount of time it took Roman women to spin, various attempts have been made to gauge this. One method is using experimental archaeology to test the amount of time required. At my own rate of spinning it would take 180 hours to spin one Roman pound of wool.@footnote{Timing myself, it took 12h 25m to spin .8 oz of wool and produce 128 yards of thread, a Roman pound is roughly 11.6 oz of wool} Based on similar experiments at the Centre for Textile Research, Mary Harlow estimates it would take roughly 900 hours to spin the 40km of thread required to make a Roman toga.@cite-footnote[Harlow2016 139] Ulrike Roth estimates that at a rate of 100 metres per hour to spin wool, it would take roughly 135 hours to produce the 13,500m of thread to weave a Roman tunic.@cite-footnote[Roth2007-Spinning 81] These experiments, however, are limited the the spinning capabilities of those conducting the tests and it is difficult to recreate the level of skill and ease that a Roman woman would have had. References to cultures who still produce thread using a @index['("spindle/whorl")]{drop spindle} are perhaps more accurate. Traditional @index['("quasillaria/spinner")]{spinners} from the Andes average roughly ninety-eight yards an hour to spin wool or alpaca fiber.@cite-footnote[Franquemont2011 13] An Incan tunic required roughly four hundred hours to spin the required thread.@cite-footnote[Bogadottir2012 51] Calculating between cultural sources, Virginia Postrel indicates that it would take 444 hours by an Andean @index['("quasillaria/spinner")]{spinner} to spin the 40km Harlow estimates for a Roman toga, 400 hours using an Indian @italic{charkha} spinning wheel, and 440 hours on a medium spinning wheel.@cite-footnote[Postrel2020 49] 


@;**TODO Chris, put in math for spinning calculations! it took me 12h 25m to spin .8 oz of wool how long would it take me to spin 11.6 oz of wool? ... I think 180

Spinning is the stage of textile production most closely associated with women. Evidence from job titles in epitaphs indicates that the job of @index['("quasillaria/spinner")]{spinner}, @italic{quasillaria}, was performed exclusively by women. Though it is entirely possible that some male domestic slaves and servants performed this task as part of their duties, there are no men commemorated as @italic{@index['("quasillaria/spinner")]{spinners}}. Furthermore, given the altogether low number of @italic{@index['("quasillaria/spinner")]{quasillariae} }mentioned in inscriptions, the activity must have been more widespread, most likely performed alongside other duties by servants and members of the household.

The precise mode of weaving in the Roman world is difficult to pin down because no Roman looms survive in the archaeological record.@cite-footnote[Wild2002 10] The warp-weighted loom was the standard type used in Ancient Greece and early Rome @figures["Amasis" "Beazley216789"] This type consists of an A-frame on either side with a beam connecting them from the top. The warp threads are suspended from this beam are weighted in clusters at the bottom relying on gravity to provide tension.@cite-footnote[Hoffman1964] The warp threads alternate between two positions, half are draped over a horizontal beam along the front of the loom with the other half hanging behind.@footnote{For the most basic tabby weave the threads would simply alternate one front, one back @cite[Lipkin2012 20]} The A-frame design gives the front of the loom an angle making it so that the threads hanging behind the horizontal beam fall straight down while those in front of the horizontal beam are further forward leaving a gap between the two sets of threads.@footnote{ A similar effect can be accomplished without the A-frame by leaning a rectangular frame against the wall without back supports.}  This gap, or shed, is the space through which the weft thread is passed. A horizontal bar called the @index['("heddle")]{heddle} connects to the warp threads which hang behind the beam. By shifting the @index['("heddle")]{heddle} from the resting position against the front beams of the loom onto the @index['("heddle jack")]{heddle jacks}, support structures which protrude from the front vertical beams of the loom, the @index['("textor/weaver")]{weaver} can shift the shed in the opposite direction, allowing the weft thread to pass through again. Since the warp threads are only secured to the loom via the top horizontal beam, the @index['("textor/weaver")]{weaver} must work from the top down. @index['("loom weight")]{Loom weights} made of clay or stone are typically the only surviving archaeological evidence of this type of loom. Signs of use on loom weights typically include vertical grooves from the hole in the weight to the top where the warp threads would have been looped through. The presence of @index['("loom weight")]{loom weights} at Roman sites suggests that this loom type continued to be used to some extent into the Imperial period. However, sets of weights large enough to supply an entire loom are scarce and certainly could not supply the quantity of fabric the population required.

A Roman two-beam upright loom consists of two upright beams on the sides with horizontal beams at the top and roughly a third of the height from the bottom. The warp threads are looped around these two horizontal beams to create tension, the thickness of the beams forms the shed.@cite-footnote[Wild1976 172] The process of weaving from this point can follow the same pattern as described above with the @index['("heddle")]{heddle} connected to the warp strands in the back. However, since the structure of this loom type is more secure with the warp threads secured at both ends of the loom, this type of loom was also more efficient for tapestry weaving. Unfortunately, this loom type does not survive in the archaeological record since it is made entirely out of perishable material and generic hardware. We know the structure of the loom from visual representations, one relief from the Forum Transitorium @figures["ForumTransitorium"] and a fresco in the Hypogeum of the Aurelii @figures["Aurelii"].@cite-footnote[Wild1976] Seneca, paraphrasing Posidonius's description of weaving, references both a loom with "threads stretched by means of hanging weights" and a loom where "the web is bound to frame," indicating that while the warp-weighted loom was somewhat old-fashioned it was still in use concurrently with the Roman two-beam loom.@footnote{Seneca, @italic{Epistulae Morales} 90.20} Since the warp threads are secured at both ends of the loom, the two beam upright loom can also be physically easier to weave on because it is possible to work from the bottom up, allowing the @index['("textor/weaver")]{weaver} to sit while weaving. In the fresco from the Hypogeum of the Aurelii, the woven portion of fabric is represented at the bottom of the loom. In the three looms represented on the Forum Transitorium reliefs, the thread and fabric are not detailed, however women sit on the ground in front of two of the looms weaving.

The amount of time required for weaving is difficult to assess because the size of the desired fabric, the style and pattern of weaving, the density of threads per inch, and the method of finishing off the ends of the fabric all impact the final time investment. Overall, however, the process of weaving on either loom type typical of the Roman Empire required somewhere between one tenth and one sixth the time investment than spinning. Karen Carr suggests that weaving a square metre of fabric on a warp weighted loom could vary from four to twelve hours per meter.@cite-footnote[Carr2000 165] Ulrike Roth estimates that it would take roughly fifteen hours to weave a Roman tunic.@cite-footnote[Roth2007-Spinning 82] Mary Harlow estimates that a single Roman toga could take two @index['("textor/weaver")]{weavers} roughly one hundred hours to weave.@cite-footnote[Harlow2016 139]

@;Add the vase depicting Penelope with her loom for a Greek example of an upright two beam tapestry loom

Both male, @italic{@index['("textor/weaver" "textrix")]{textor,} and female, @italic{textrix,} weavers} are identified by job titles in epitaphs.@cite-footnote[LarssonLoven1998 75] Though professional @index['("textor/weaver")]{weavers} are more frequently men, weaving within the household is associated with women. In his description of the mythological weaving battle between Arachne and Minerva, Ovid describes the dressing of the loom and the process of weaving:

 @nested[#:style 'inset]{And both, at once, selected their positions, stretched their webs with finest warp,
  and separated warp with sley. The woof was next inserted in the web by means
  of the sharp shuttles, which their nimble fingers pushed along, so drawn within
  the warp, and so the teeth notched in the moving sley might strike them.—Both,
  in haste, girded their garments to their breasts and moved their skillful arms,
  beguiling their fatigue in eager action.@footnote{ While somewhat unclear, the Latin word @italic{harundo} (a reed or cane), translated here as 'sley', likely refers to the @index['("heddle")]{heddle}. Ovid, @italic{Metamorphoses} 6.1.55}}

@index['("Dye Shop")]{Dyeing} of textiles could be done at nearly any stage of textile production, raw wool, spun fibers, or woven fabrics.@footnote{Diocletian's Price Edict lists prices for both unprocessed silk and wool as well as spun silk and wool dyed purple Diocletian, @italic{Edictum de pretiis rerum venalium} XXIV.1, if this was done at these stages for purple dye, was presumably done at various stages in other colors as well. @cite[Wild1976 168]} Furthermore, used fabrics were often re-dyed to refresh their appearance. @index['("Dye Shop")]{Dyeing} shops can be identified archaeologically by deep lead cauldrons with furnace installations below to heat the chemicals.@cite-footnote[Flohr2013 60]

@index['("fullery/fullonica")]{Fulling} is the final stage of Roman textile production. After the cloth is woven it is exposed to heated liquid, agitated, and combed to tease out some of the fibers. After the fulled fabric dries it shrinks down in size tightening the weave. In essence, the combination of heat and friction effectively felts a layer on top of the cloth. The final step is to finish the cloth by sheering off any excess fibers to create a smooth surface.@cite-footnote[Wild2002 22] Fulling requires specialized equipment installed into a workspace and therefore it is easily identifiable in the archaeological record. A @index['("fullery/fullonica")]{fullery} is typically divided into stalls with basins sunken into the floor separated by partial walls with waterproofed coating.@cite-footnote[Flohr2013 62] 

While it is unclear where the distinction between domestic and commercial production of textiles begins, or even if that distinction has any real value, some stages of production could be done anywhere while others required more specialized spaces. The processing of the wool, spinning, and weaving described above could be performed within a domestic space, and archaeological finds support this. The large cauldrons, basins, and vats required for @index['("Dye Shop")]{dyeing} and @index['("fullery/fullonica")]{fulling} necessitated facilities that multiple households and commercial endeavors would have shared. After the production stages, the finished products were either used within the household where they were produced or fed into the commercial textile industry where they were packaged, shipped, and sold. According to Roman practice, clothing was typically woven as a single piece on the loom, so little alteration or tailoring was necessary from weaver to consumer.@cite-footnote[Wild2002 22] Job titles for the textile manufacture and commerce which happened outside of the house including @index['("Dye Shop")]{dyers}, @index['("fullery/fullonica")]{fullers}, and merchants were almost exclusively associated with men. 


 @section{Methodology}

@; **Note: Due to the current crisis and the closures of libraries, I have not yet expanded and updated the methodology and literature reviews since I wrote my prospectus. For the time being, I've left them here for reference.

 @; CW: I'm unsure what the citation for Hedon2007 would be about... that it is "important" to do that?  I wonder if you can back it up with something specifically said by the author, since this sounds like a subjective claim.  Likewise with Meskell2001.

For this research, I will chiefly employ the methodologies of social archaeology and social history. These two approaches varry primarily in their source material; however, given the limited amount of evidence for textile production overall, I incorporate both archaeological and written sources into my evidence. It is important to examine the social and cultural factors that inform textile production and the people involved in performing it.@cite-footnote[Hendon2007] In order to understand the extent of social and cultural relationships between these women and their craft, it is important to look at multiple aspects of identity. Because the category of 'women in the Roman empire' is far too broad on its own, I will consider how factors such as status, class, and age effect this discussion.@cite-footnote[Meskell2001] It is likewise important to note that these factors can change over a woman's lifetime and in relation to the social situation.

Since this research focuses largely on the social history of women in particular, it will also be greatly informed by feminist methodologies of historical and art historical research. While the early stages of feminist -history, -art history, and -archaeology focused on the gaps in our understanding of women's roles in history and writing women back in, this work has largely been accomplished over the last fifty years of scholarship. My research focuses on recontextualizing women's roles in textile production from a more intersectional approach and analyzing how the biases of antiquity in regards to gender, slavery, and labor overall have been perpetuated in scholarship to categorize women's labor as domestic while maintaining relative silence on women's roles in commercial endeavors. As Amy Richlin highlights, locating women in the Roman world involves arguments with silence to counteract the lack of women's voices.@cite-footnote[Richlin2014]


I explore both domestic and commercial production of textiles and the various ways that those methods of production overlap. In order to accomplish this, I also incorporate economic history into my approach. This methodology assesses textile production within the larger commercial structure of the Roman Empire. Scholarship in ancient economic history since 1973 has hinged on the watershed publication of Moses I. Finley's @italic{The Ancient Economy}.@cite-footnote[Finley1999] This work upended the notion within the research of classical antiquity that economics of the ancient world could be understood as in terms of the same economics we are biased from within our current era and led to great interest in perceiving the ancient world in terms of the study of issues of historical justice. Nonetheless, other historians such as Paul Cartledge have bemoaned that this has lead to research surrounding ancient economics being split into two areas: that of the study of raw data, material culture, and archaeological findings (modernists), and that of a strand of social analysis which, while trying to avoid recognized contemporary social biases (primitivists), which -- while useful for upending stale assumptions -- may be injecting its own cultural biases within the vacuum that remains.@cite-footnote[Cartledge2012] Jean Andreau in particular has pointed out that the field has amended significantly from this dual-state, and that in particular much progress was made outside of English-speaking circles.@cite-footnote[Andreau2012] My own work does not fall cleanly on one side of this supposed duality. My interpretation of the Roman textile economy acknowledges contemporary biases that lead to a lack of account for the roles that women played in that larger social structure; no doubt this sounds extremely primitivist.  However, I have tried to ground the shape of my claims within data, archaeological findings, and material culture in a way that has some modernist flavor. However, the scope of this work is to account for women's roles in textile production, not to form a concrete understanding of the Roman economy as a whole.@footnote{ It is of no little significance to this research that these economic histories often omit women entirely from their analysis. Of the mere six times that Finley mentions women in @italic{The Ancient Economy,} four are in the context of slavery.} The particular economic model that I focus on is cottage industry, a system in which goods are produced within the home with the intention of feeding into a larger market economy.@footnote{A more detailed discussion of cottage industry systems will follow in chapter 6. See @cite[Boeke1942], @cite[Boeke1953] and @cite[Prentice1983]} This model allows for the permeability between domestic and commercial production that the Roman textile evidence suggests. 


For this study, I employ an interdisciplinary approach to studying artifacts through the inter-related fields of archaeology and material culture which both use material objects as evidence of past societies. Archaeological evidence from my three case study sites of Karanis, Trier, and Ephesus as well as other sites across the Roman empire forms a central part of my arguments. As Jules David Prown highlights, "material culture is thus an object-based branch of cultural anthropology or cultural history."@cite-footnote[Prown1993 1] It therefore incorporates aspects of the above methodologies as well as incorporates issues of consumerism and the materiality of objects.@footnote{ For more comprehensive histories of material culture as a field of study see @cite[Buchli2004] and @cite[MartinGarrison1997]} Within this framework, we can also approach the use and re-use of artifacts through object biographies or life-cycles, allowing for shifting significance of an object given the context of its use.@cite-footnote[Dannehl2009 124]

In conjunction with this research, I have learned the basic skills of combing, spinning, and weaving wool on a warp weighted loom and a two-beam upright loom @figures["BlueWoolSpindle" "TabletBand" "Loom" "Shed"]. I constructed a small-scale loom (24 inches wide) based on ancient sources and reconstructions from textile historians such as Eva Hoffman, Elizabeth Barber, and John Peter Wild.@footnote{ @cite[Hoffman1964]; @cite[Barber1994]; @cite[Wild1976]} I have also constructed a full-sized Roman 2-beam loom primarily from the visual evidence from the Forum Transitorium and the tomb of the Aurellii @figures["ForumTransitorium" "Aurelii"] This experiential knowledge of the technologies used has helped my understanding of the material. After honing these skills, I was then able to collect quantitative data on the amount of time it took to spin a pound of wool, the length and thickness of the resulting thread, and the amount of material loss.@footnote{This data is cross referenced with data on wages and cost of materials recorded in Diocletian's Price Edict in Chapter 6.} To this extent, I have practiced experimental archaeology, the attempt to lend credence to archaeological interpretations through experimenting with ancient technologies and crafts. Experiential knowledge of crafts can lend us opportunities to revisit assumptions we may derive from purely reading primary and secondary sources. For example, Elisabeth Trinkl asserts that decorative @index['("distaff")]{distaffs} from Ephesus were not used because they do not show signs of wear and have too small a space for wool to be wrapped around them to be functional.@cite-footnote[Trinkl2008] This is a reasonable interpretation to make. However, based on my own experience spinning, it seems unlikely that fluffy wool roving would leave signs of wear on bone. Furthermore, I have personally used a @index['("distaff")]{distaff} with a smaller surface area.@footnote{I am unsure whether Trinkl has any experiential knowledge of spinning, and do not mean to imply that she does not, though I didn't see any mention of an experiential approach in her research; I am merely sharing my insights from my own experiental crafting experience. I have further plans to test this assumption via the use of reproduction @index['("distaff")]{distaffs} using clay, but the current state of self-isolation brought on by the COVID-19 pandemic means I do not have access to the tools and materials to complete the project}

@;ToDo: Experiental/Experimental Archaeology Appendix: and will record myself demonstrating spinning and weaving on the two types of loom. I will probably write up a brief apendix on my experiences with experimental archaeology for my dissertation research.

When well done, experimental archaeology can allow scholars to cast projections on historical output. For example, Ulrike Roth gave a range of projections for the average output of a rural Villa, giving three potential levels of economic potential depending on the number of workers, looms, and productive hours in a day.@cite-footnote[Roth2007-Spinning 82] There are, however, limitations to the conclusions we can make about the time and output of ancient craftsmen.@footnote{@cite[Lipkin2012]; @cite[Roth2011]} There is no way for a scholar to simulate the time investment or early and constant training of a lifelong skill that ancient craftswomen would have gained. These techniques are more valuable when paired with ethnographic study of cultures that continue to use similar technologies.

@section{State of current research}

 While this study covers little Greek material, a scholarly foundation of research regarding the relationship between women and wool working in the Greek world has been crucial to framing my research. Most notably, Elizabeth Barber efficiently breaks down the central role that women played in the early development of textile production in her 1994 book @italic{Women's Work}.@cite-footnote[Barber1994] Furthermore, studies on the dispersal of spinning and weaving tools have been frequently used to either locate or disqualify the notion of women's spaces in the Greek house.@footnote{@cite[Cahill2002]; @cite[Nevett1995]} These discussions of gendered artifacts in domestic contexts can be useful templates for Roman domestic sites as well.



 @; NDC: Is this important? Or do the native Italic tribes have their own strong weaving traditions? Your discussion below focuses on the later.

Some attention must be given to cultural contact between Greeks and native Italic tribes where direct social interaction occurred due to Greek colonies in Italy and Sicily. Extensive evidence for trade between the Greeks and Italic cultures survives including many Greek vases discovered from Etruscan tombs.@cite-footnote[Bundrick2012] Margareta Gleba has written extensively on textile production in Pre-Roman Italy and edited two volumes on the topic.@cite-footnote[Gleba2009]  Marianne Kleibrink, Jan Kindberg Jacobsen and Søren Handberg discuss the merging of Enotrian and Greek religious practices resulting in a cult, likely to Athena, associated with wool production.@cite-footnote[Kleibrink2004]  Lin Foxhall and Alessandro Quercia have written on women's production and women's networks in Italic communities.@cite-footnote[Quercia-Foxhall2015] Sana Lipkin analyzes evidence of textile tools in Central Tyrrhenian Italy from Pre-Roman through the Republican period. She breaks down the types of fibers used, the types of tools used, and the cultural implications about women. Her evidence is primarily archaeological backed up with literary and epigraphic sources.@cite-footnote[Lipkin2012]

 @;Talk more about Miko Flohr and Jongman who have written extensively about the textile trade networks of Rome
Research on textile production in the Roman Empire has primarily focused on commercial production and trade.@cite-footnote[Jones1960] Willem Jongman and Miko Flohr have focused on the Roman textile industry largely through analysis of @index['("fullery/fullonica")]{fulleries} and @index['("Dye Shop")]{dyeshops}.@footnote{ @cite[Jongman1988]; @cite[Jongman1997]; @cite[Jongman2000]; @cite[Flohr2013]; @cite[Flohr2013Fullo]; @cite[Flohr2016]  } The most prolific scholar of Roman textiles, J.P. Wild, focuses primarily on production techniques, analysis of surviving textile fragments, and trade.@footnote{@cite[Wild2002]; @cite[Wild1999]; @cite[Wild1976]; @cite[Wild1970]; @cite[Wild1976-Gynaecea]} Lena Larsson Lovén@footnote{@cite[LarssonLoven2007]; @cite[LarssonLoven2016]; @cite[LarssonLoven1998]; @cite[LarssonLoven1997]} and Suzanne Dixon@cite-footnote[Dixon2000] have approached the issue of women's roles in textile production in the Roman world and both focus on wool work as iconic of feminine virtue in a society whose textile production is primarily commercial. Both Larsson Lovén and Dixon use funerary epigraphic and iconographic sources but do not include direct archaeological evidence in their analyses. Daniella Cottica discusses the relationship between women and wool working in Rome including the deposition of @index['("spindle/whorl")]{spindles} as grave goods;@cite-footnote[Cottica2006] however, she does not have the space to fully analyze these materials in her brief articles. Penelope Allison has included weaving implements in her analysis of the material culture of Pompeiian households, yet the small quantities of these objects is not conclusive as to the extent of domestic production of textiles.@cite-footnote[Allison2006]  @index['("loom weight")]{Loom weights}, for example, were frequently found in Pompeiian houses, yet rarely in large enough quantities to supply a loom.

@; NDC: Conclusions? what does this tell you? It's very interesting to me -- beyond the scope of this dissertation, but well worth thinking about.

 @; NDC: you mention at least three works in the text; what are they, the authored articles and two edited vols?

Therefore, while the various aspects of this topic have been well discussed in the scholarly literature, there have been few attempts to apply a holistic approach to women and textile production in the Roman Empire. Daniella Cottica's “The Symbolism of Spinning in Classical Art and Society” rather deftly analyzes this topic thematically utilizing both Greek and Roman evidence; however, the breadth of the topic exceeds the bounds of an article spanning only 20 pages including copious images.@cite-footnote[Cottica2006] Lena Larsson Lovén's work largely covers the symbolic aspect of this topic through literary, epigraphic, and iconographic sources. Since most of her evidence is funerary, she does not engage with the way this related to the lived experiences of Roman women. Sana Lipkin incorporates evidence from domestic, religious, and funerary contexts to discuss the relationship of women to textile production. Yet, she does not discuss the ways that the association between women and textile production continued through the changing social, political, and commercial environment of the Roman Empire because she uses the shift from domestic to commercial production in the Imperial period as a turning point and ends her analysis with the Republican era.@cite-footnote[Lipkin2012 12]  All of the aforementioned research is of excellent quality and invaluable to this study; the distinction for this dissertation is of focus and scope. 

@; NDC: So do you extend SL's work into the Imperial period?

My contribution to the field will be in cross-analyzing the evidence of textile production from three sites -- Karanis, Trier, and Ephesus -- with evidence evidence from the Roman Empire more broadly to examine how local textile production and the associations between women and textile production changed or persisted as the Roman empire expanded.  In this way, is a continuation of the legacy of Elizabeth Barber and Sana Lipkin, extending both the analysis of the existing surveyed timescales and bringing the research forward into the Imperial period.  This dissertation takes into account archaeological, epigraphic, iconographic, and literary sources for a holistic approach to understanding how the evidence reflects the cultural realities of Roman women or invokes wool working as iconic of feminine domestic economy.  Additionally, I will look at the agency of women in textile production, particularly how women's labor translates into economic potential for both women individually, for their households, and as an essential component of the textile economy as a whole.  The model of a cottage industry illuminates how this economic potential need not be divorced from well-documented domestic production.

There is a running theme throughout this dissertation: sometimes the hole that delineates the absence of evidence is itself the shape of the evidence.  Given the relative dearth of evidence for women's agency in the Roman world overall, the lack of women's perspectives in the written sources, and the gender biases of the past and present, my analysis will hinge on reading the empty spaces between the evidence available.@cite-footnote[Richlin2014 1]  


 @;section{Conclusions}

 @; - things are forthcoming from research
 @; - however much is known
 @; - significant physical evidence due to climate and desertification
 @; - potentially, there is evidence that the symbolic value of women performing spinning is represented through iconography blah blah...
 @; - we have good evidence that commercial production occurred here, which leaves us to ask, "who is doing the actual production?"
 @; - contracts give us proof that at least in some cases women were involved in commercial production
 @; - but there is also good reason to believe that women would be responsible, since they are responsible in domestic spaces


 @make-footnote[]

 @(generate-chapter-bibliography)