*************************************
* Spinning Experiment - Readme File *
*************************************

The Spinning Experiment took place during the Textile Forum at Eindhoven in September 2009. 

An article with an analysis of the Spinning Experiment will be published as "Determining Influences on Yarn spun on a Hand-Spindle – Results of a Large-Scale Spinning Experiment" in the "Archäologie Österreichs Spezial" containing the conference papers of the 2010 OEGUF conference about experimental archaeology. 

Spinning time was divided into one-hour blocks, each with a different combination of spindle (one of five experiment spindles) and fibre (one of two different and differently prepared wools). On five consecutive days, two of these combinations were used for spinning, the first one always with Merino wool, the second one with Bergschaf wool. This division was primarily made not to overtax the participants; spinning is a physical activity, and modern hobby spinners are not always spinning for a prolonged time in one session. Two hours of spinning time were supposed to be achievable by all participants without getting influences due to fatigue.

Spinners were asked not to turn the experiment into a competition; they were bid to concentrate on the spinning, but allowed to chat; thus group dynamics of course come into play. There was noticeably less chatting just after start of a session, while spinners got acquainted with their tools, and throughout the whole sessions with the spindles rated as "difficult" or "hard to handle" by the spinners; thus more concentration seemed to be spent on spinning with these spindles than other, "easier" ones. Of course, chat intensity and volume level also differed depending on the topic. Occasionally, when discussions got lively or people explained something, there were short pauses in the spinning process. These were however infrequent and short enough to probably not have too much influence on the final result.

Spinner K had to leave due to personal reasons after the 2nd day, thus K has been left out of the final analysis of the experiment. Her data (the first four sets) are included in the table, but no visual survey cards were made from her threads yet.

Weather has possibly also had an influence, since spinning took place out of doors under a tent roof. Weather conditions were roughly documented in one of the tables, and comments on weather and temperature were also made by the spinners in the questionnaires.
 
It resulted in a dataset of 14 spinners with ten hours of spinning time each; the raw data that came out of the experiment can be found in the database.

You are welcome to use the data collected at the Spinning Experiment for your own evaluations, and I would be very happy to hear about your conclusions regarding the data. 

Please be aware at all times that spinning with a hand-spindle is not yielding a totally homogenous thread as modern machines do. There are always variations in the thread; you can see this on the visual survey cards. In addition to this, the outdoor conditions can have had an influence. Weighing equipment accuracy is also limited, with 0.1 g steps of the digital scales. Please do not see the numbers as an exact documentation of the spinner's threads, but as an average number. 

_all_data contains the complete dataset. In some cases, spinner's comments were longer than fit into the database field; you find the rest of the comments in the file named long_comments.txt.

Techniques and equipment used for documentation and analysis:

Weight measurements were made with a Soehnle Ultra 2.0 digital scale, accuracy 0.1 grams. Scales were calibrated using the 200 g calibration weight before each session analysis. Photos were taken with a Canon 450D. Yarn length measurement was made on a high-precision reel in the Textiles Prüfamt Münchberg, Germany; lengths were measured to a 1 cm accuracy. For the re-spooling before making the survey cards, a simple length measurement tool (made by Woll-Wolff) was used; its display accuracy is one metre.

Spinners were photographed after about half an hour of spinning time and again after 50 minutes of spinning time. Spinner photographs are available upon request, please email to katrin.kania@pallia.net. The filled spindles were photographed after the session, then the yarn was transferred to a plastic drinking straw; these yarn cops were photographed again. Please contact me if you would like photos of the yarn cops.

After measuring total length on the precision reel, 2/3 of each yarn length were transferred to a paper core, with the starting thread innermost. Lenghts were calculated and the spooling length was measured to the approximate half metre. The remaining yarn, the last third of each spinner's output,was used to make a visual survey card. In addition, the approximate yarn diameter was measured by wrapping the yarn around a wooden dowel on a length of 3 cm and then counting the number of wraps. Wrap counting was made with the thread remaining after filling the survey card for spinners E, G, I and O; for the rest of the spinners, the end of the yarn before wrapping the survey card was used.
For the samples too short for filling the survey card with the last third of the spinner's output (equalling less than 4 metres), the survey card was filled with the center 4 metres of the sample. Wrap count in these cases was taken with the first part of the yarn used for the survey card.

Wrap counts for all spinners except E, G, I and O were taken by Katharina Gottschlig.

The visual survey cards were scanned in as an easy way to digitise them; you find them labeled with Spinner and Session number, spindle ID and a remark "start at xx of yy m"; the latter gives the approximate starting point of the visual survey card wrapping of the total length of the sample.