Keywords

1 Introduction

Hanging scrolls are among the most recognized of the many forms of binding and displaying Japanese paintings and calligraphy (Fig. 1).

Fig. 1.
figure 1

A scroll hanging in an alcove

Full size image

Scrolls are unique in that they are hung on a wall or in an alcove only during viewing, and stored in a box otherwise. When the scrolls are rolled up for storage, the artwork faces inside, protecting it from exposure to light and air to minimize damage. This binding method is garnering attention even within the area of preserving cultural assets for its ability to preserve artwork in the long-term.

The scroll has to be able to roll up smoothly without causing any creases. And, when hung for viewing, the scroll has to hang straight without curling up. To enable this, the back of the scroll is typically lined with 3–4 layers of Japanese washi paper. The washi is attached to the back with the wheat starch. The second layer of washi and beyond, in particular, requires wheat starch with weak adhesive to keep the backing from hardening after it’s dried. This adhesive is what makes the smooth and creaseless scrolling possible. However, in order to withstand frequent handling, the washi needs sufficient adhesion. This is where the surface of the washi is pounded down with a brush to enhance adhesion (Fig. 2). This technique is called pounding.

Fig. 2.
figure 2

Pounding in action

Full size image

Pounding with too much force can damage the surface of the backing, but also potentially the artwork that lies several layers beneath. Pounding with too little force, on the other hand, makes for ineffective adhesion.

This pounding technique requires repetitive tapping with the right amount of force evenly across a wide area. The acquisition of this technique is said to require a lot of training. Much like other traditional Japanese techniques with a high degree of difficulty, “watch and learn” is currently considered the main correct method of acquiring this skill.

For this research, we had two subjects, one an expert and another a non-expert, perform the pounding technique with their right arm, which we hooked up to an electromyograph to study the muscle activities involved with the work. We hope the results of this research will deepen our understanding of the pounding technique, a skill that takes a long time to learn, and help contribute to the acquisition of this skill.

2 Structure and Material of the Hanging Scroll

Hanging scrolls consist of several layers, starting with the main sheet, which would be a painting or calligraphy, followed by the first, second, third and the final linings. The washi paper used for each layer differs slightly in terms of where and how they were made. Each washi layer contributes to the aforementioned rolling and unfurling possible, providing the scroll with sufficient strength. To attach the first lining, wheat starch is turned into paste and used as adhesive. The wheat starch is heated in a large pot for about 50 min, then cooled off naturally overnight to produce wheat starch glue with strong adhesion. This is used to attach the first layer of washi to the main sheet. This first sheet serves as the main sheet’s backing. However, using this adhesive for the second lining and beyond will only harden the layer more than necessary after the starch dries and prevent the scroll from rolling and unfurling smoothly. Which is why, for the second layer and beyond, an adhesive referred to as aged glue, with a weak adhesive strength, is used. To make aged glue, wheat starch powder is heated, turned into paste, then poured into a large jar and sealed with a wooden lid. After sitting in a cool and dark place for 10 years, the glue is ready. Aged glue has low adhesive strength compared with regular wheat starch glue. Long-term storage causes the starch to deteriorate and lose molecular weight, and because the microorganisms metabolize enzymes in lower volume compared with regular wheat starch glue, aged glue has weaker bonding power. Figure 3 shows the cross-sectional view of a hanging scroll.

Fig. 3.
figure 3

Abbreviated diagram of the cross-section of a hanging scroll

Full size image

3 Experimental Method

3.1 Subjects

The electromyograph experiment was performed on a 22-year expert and a 6-year non-expert. Details about each subject are shown in Table 1. The expert is acclaimed for his ability to deliver a quality product with his pounding technique. The non-expert, on the other hand, had received regular warning from the expert about pounding too strongly, and is yet to receive permission to begin pounding work. Furthermore, during an interview conducted before the experiment, we learned that the expert can perform pounding work continuously for more than several hours without feeling tired, whereas the non-expert begins to feel fatigue after 30 min of continuous pounding.

Table 1. Data about the subjects
Full size table

3.2 Method of Experiment

The subjects were asked to begin pounding work after gluing on a second lining on a plain sheet of silk measuring 40 square cms already backed with a first lining. Each subject’s muscular movements will be measured in the process. The pounding action is to begin at front right, moving to the left in even and continuous motion to adhesion enhancement. The number of times the backing is pounded and the amount of time spent pounding differ by subject. Which is why, we measured the muscular movements of just the right arm. The parts we measured are shown in Fig. 4. There are 9 locations, which are listed below.

Fig. 4.
figure 4

Locations that were tracked.

Full size image

Abductor pollicis brevis, Deltoid middle, Upper trapezius, Middle trapezius, Biceps brachii, Triceps brachii, Extensor carpi radialis brevis, Flexor carpi radialis, Interossei dorsales.

Electrodes (Nihon Kohden, web-1000) that were attached to the subjects’ arms measured muscular activity while they performed their pounding work in an environment similar to the one they usually work in. The sampling rate was set at 1000 Hz. Two cameras – one in the front and another to the right – captured them at work.

4 Measurement Results

Results of each subject’s muscle activity in the 9 locations are shown. From the overall results, we will look at which muscles were being used during the pounding action, then pinpoint which muscles were used per one pounding motion for each of the 9 muscles.

4.1 Muscles that Worked During Pounding

Figure 5 shows the expert’s muscle activity as expressed in raw waveforof the EMG; Fig. 6, that of the non-expert. Of the 9 locations monitored, the biggest difference between the expert and non-expert was in their abductor pollicis brevis activities. On the expert’s graph, a big wave could be seen immediately following his movements, but after that, there were no large waves until after the work was completed. The big wave created by the abductor pollicis brevis immediately after monitoring begins relates to the expert’s movements before pounding work is about to begin – a wave generated by the gripping action his abductor pollicis brevis makes when re-adjusting the pounding brush in his hands. On the other hand, the non-expert’s abductor pollicis brevis creates big waves throughout the monitoring period that becomes more dramatic especially in the last-half of the pounding process. This indicates the non-expert is using force with his thumbs during the work to stabilize his grip.

Fig. 5.
figure 5

Results of the rest on the expert

Full size image
Fig. 6.
figure 6

Results of the test of the non-expert

Full size image

We were able to confirm waves from the expert’s chart in the following muscle groups: the middle trapezius, extensor carpi radialis brevis, and interossei dorsales. But when juxtaposed with the video, for most the muscle movements, we found that the biggest waves were formed right after the pounding, when the pounding brush was being lifted. Furthermore, we were not able to detect any regular waves in the triceps brachii.

As for the non-expert, we detected large waves occurring in most muscle groups, compared with the expert in areas other than the aforementioned abductor pollicis brevis. Also, the waves tended to get larger toward the last half of the process. When matched up against the video, we were able to confirm the occurrence of large waves right after the pounding motion, which is much like the expert, but also when he is bringing the pounding brush down. We confirmed that the non-expert was consistently using all 9 monitored muscles to a greater degree than the expert.

4.2 Activity Per Muscle Group of the Expert and Non-expert

Here, we can find the type of activity that takes place in each muscle group for each pounding motion by smoothing the data.

Abductor Pollicis Brevis.

The abductor pollicis brevis is activated during the palmar abduction of the thumb to grasp an object. It is also a functional position used by the expert. This muscle attaches itself to the scaphoid bone and the retinaculum of flexor, to the sesamoid bone, abductor proximalis. The graph shows that the expert uses very little of this muscle, which makes it clear that he does not hold the brush firmly in his hands. Furthermore, we can say that he pounds without creating tension in his arms. If anything, he is relaxed. His actions are well timed without any waste. Because of this, the timing of the pounding, the force he applies to the brush and even the intervals are well balanced. For the non-expert, on the other hand, high muscle activity and use of first joint can be seen. This is most likely due to his firm grip on the brush, which he holds with greater force than his expert counterpart. This is shown in Fig. 7.

Fig. 7.
figure 7

Abductor Pollicis Brevis

Full size image

Deltoid Middle.

The deltoid originates at the lateral side of the clavicle, the acromion, and the spine of the scapula,and attaches to the deltoid tuberosity. Its function is to abduct the shoulders. The middle fibers of the deltoid, in particular, extends from the acromion to the deltoid tuberosity and is the main muscle to help the shoulders abduct. When compared against the non-expert, the expert’s muscle activity while pounding tended to be long and intense. This is because the expert, more so than the non-expert, uses his middle deltoid fibers to keep the arm slightly abducted while working. We believe this action contributes to stabilization of the shoulder joins. On the other hand, the non-expert was considered incapable of stabilizing his shoulder joint and continuing his work. Furthermore, we were able to see that the non-expert’s upper arm was actively lifted at about the same time that he was pounding with the brush. This means increased muscle activity was likely due to the shoulder abduction while he was lifting his arm. This is shown in Fig. 8.

Fig. 8.
figure 8

Deltoid middle

Full size image

Upper Trapezius.

The upper trapezius connects from the lateral end of the clavicle, the acromion and the spine of scapula to the occipital external protuberance, nuchal ligament and the spinous processes of C1-C7. This muscle is responsible for preventing scapular depression. Between the expert and non-expert, muscle activity is overwhelmingly high but for a short duration for the expert. The non-expert, on the other hand, differs from the expert in that his muscle activity is low, but for a longer duration. This indicates that one action of an expert is completed in a shorter amount of time, and muscle activity per action is more intense. This is because when moving the brush up and down, the expert fully engages his upper trapezius when lifting his arm, stabilizing the surface of his shoulder blades. This movement with the brush is called pounding, but more accurately, the moment the brush hits the target, the expert lifts his arm to move on to the next action. The non-expert’s data showed less action in this movement, which was insufficient to stabilize the shoulder blades, unlike the expert, and resulted in a difference in the quality of the work. When performed by the expert, the lifting action in the arm was swift, while the non-expert barely raised his arm – all of which became evident after smoothing the data. This is shown in Fig. 9.

Fig. 9.
figure 9

Upper trapezius

Full size image

Middle Trapezius.

The middle fibers of the trapezius muscle are connected from the upper portion to the middle of the spine of the scapula and adducts the shoulder blades. When comparing activity in this muscle between the expert and non-expert, data shows a very high level of prolonged activity for the non-expert, whereas the opposite was true for the non-expert in both activity and duration. It became clear that when working, the expert actively engages the middle fibers of the trapezius without adducting them. This finding was unexpected. Ordinarily, when performing an action, the shoulder blades need to be stabilized, which is why, with regards to the expert, we were anticipating that the data would show an active engagement of the middle fibers of the trapezius. But the results indicated otherwise, showing low key activity in that area. This points to the fact that while working, the expert doesn’t allow his shoulder blades to adduct. This was evident even when observing his movements. The shoulder blades were rather in a state of abduction, and upward rotation, enabling him to work with just a moderate amount of muscle activity. We believe the non-expert adducted his shoulder blades and over-exerted while working. This is shown in Fig. 10.

Fig. 10.
figure 10

Middle trapezius

Full size image

Biceps Brachii.

The two-headed biceps brachii originate at two locations: the supraglenoid tubercle and the coracoid process of the scapula, and insert into the radial tuberosity and the bicipital aponeurosis to the fascia on the medial side of the forearm. They flex the elbow and the shoulders and supinate the forearm. With the expert, the movements were bimodal, indicating an increase in the start of the activity, while the non-expert’s results were trimodal, showing an increase in bimodal activity. Activity was much less and shorter compared with the non-expert’s. This means the expert’s movements were shorter than the non-expert’s. By bringing the pounding brush down and immediately preparing to lift it, the expert’s muscle activity increases in an instant, showing that it is an instantaneous movement. However, with the non-expert, the same activity takes place longer and with more intensity. And the timing of the movements are slightly slow. This is because the non-expert, rather than lifting the brush, potentially moved his biceps at the same time as his triceps without moving his elbow joint. This shows that while the movements of the expert and the non-expert may appear similar, data shows that their movements are completely different. This is shown in Fig. 11.

Fig. 11.
figure 11

Biceps brachii

Full size image

Triceps Brachii.

The triceps originate in the infraglenoid tubercle and attaches to the olecranon. They are responsible for extending the elbow and, when activated together with the biceps brachii, fixate the elbow joint. They also aid secondarily in fixating the shoulder joint. Results were bimodal for both the expert and non-expert. With the expert, muscle activity was light and short. For the non-expert, muscle activity was high and very long. This showed that the non-expert was using force to pound the brush on the washi paper, while keeping his biceps and elbow joints fixated. Activity was low-key for the expert, and what the bimodal indicated was that compared with the non-expert, he didn’t exert much force and used the least amount of muscularity. Furthermore, the bimodal results also showed that in the picture of the overall movement, he extended his elbow, and in the last half, flexing the elbow served as a sort of brake on the action. Results for the non-expert were bimodal like those of the expert, but the timing differential with the biceps brachii activated the triceps, then the biceps after that, followed by both the triceps and the biceps. Results indicated that the non-expert’s biceps and triceps were working in excess, compared with the expert. See Fig. 12.

Fig. 12.
figure 12

Triceps

Full size image

Extensor Carpi Radialis Brevis.

For this research, we focused on the extensor carpi radialis brevis and the flexor carpi radialis muscles as the main muscles of the forearm. The extensor carpi radialis brevis originates at the lateral epicondyle of humerus and inserts into the posterior base of the third metacarpal. These are muscles that extend and abduct the hands at the wrists. Furthermore, they work together with the palm muscles to help stabilize the wrist. This study showed unimodal results for both the expert and the non-expert. The expert’s activity in this muscle was comparable to that of the non-expert, but the duration of his muscle activity was considerably shorter. We can see from this study that he was dorsally extending and radially flexing his wrist instantaneously. Viewing the timing of the expert’s movements, activity heightens in the latter half. We can see from this that he engages his extensor carpi radialis brevis to lift the pounding brush for an instant after pressing down into it. The non-expert, however, that muscle is continuously activated, resulting in a difference in work quality. This is showin in Fig. 13.

Fig. 13.
figure 13

Extensor carpi radialis brevis

Full size image

Flexor Carpi Radialis

One can say that muscle activity involving the flexor carpi radialis and the extensor carpi radialis brevis defines the muscle activity of the forearm. They do not indicate independent muscle movement, but rather the flexing and extending activities of the forearm. The muscle originates on the medial epicondyle of the humerus and inserts on the anterior aspect of the base of the second metacarpal. The flex and abduct the wrist and help stabilize the joint. The extension of the flexor carpi radialis can contribute to the radial flexion of the wrist. The results are bimodal for both the expert and the non-expert, but the non-expert’s muscle activity in the latter part is low. Engagement in the flexor carpi radialis triggered a volar flexion in the wrist, causing a reflexive movement in the extensor carpi radialis brevis that resulted in the instantaneous dorsal and radial flexion of the wrist. In order to put a brake on the wrist movement, it could be that the latter-half activation of the flexor carpi radialis is preventing a dorsal extension in the wrist. The non-expert is also engaging the dorsiflexor, which is overly fixating the wrist joint, potentially causing overexertion in the arm. See Fig. 14.

Fig. 14.
figure 14

Flexor carpi radialis

Full size image

First Dorsal Interosseous Muscle.

We’ve called attention to the first dorsal interosseous muscle and the abductor pollicis brevis as the extrinsic hand muscles. The first dorsal interosseous originates at the the first metacarpal and the radial side of the second metacarpal bone and inserts into the base of the second proximal phalanx. The first dorsal interosseous assists in thumb adduction and is activated when gasping. Research results differed for the two subjects, with the expert’s data being bimodal, unlike the non-expert’s. When comparing the timing, the muscles were engaged when the brush was being pounded. This indicates that the subjects were gripping their brushes while pounding. Furthermore, the expert’s muscle activity in the latter half is characterized by a momentary lift in the brush immediately after pounding, then after that activity to stop the brush from bouncing up. Data from the non-expert shows no such bimodal characteristics, showing mostly engagement in the first half. This is shown in Fig. 15.

Fig. 15.
figure 15

First dorsal interosseous muscle

Full size image

5 Conclusion

For our research, we smoothed out data from the electromyograph to extract the characteristics of each muscle activity. The result was that overall, the amount and duration of muscle activity of the non-expert was considerably higher than those of the expert. Data smoothing revealed that this caused excess tension in the non-expert’s arm while he pounded. It was consistent with the non-expert’s statement during the interview portion of the research that he was exhausted after just a short amount of time. By contrast, the expert rarely used force, and was relaxed through the shoulders. The only muscle the expert engaged to a greater degree and for a longer duration than the non-expert was the middle fibers of the deltoid middle. We interpreted this to mean that the deltoid abducted the shoulder joints and held them in that position throughout the duration of the action. Then, the trapezius would depress momentarily from the first-half, but rise up again after that to form a trimodal pattern. In order to move the shoulder, elbow and wrist joints rhythmically the middle section of the shoulder joints have to be stable, which show up in the results of this study. Additionally, a dramatic increase in activity was seen in the expert’s upper fibers of the trapezius which, seen that it was accompanied by a sudden lift in the pounding brush, indicates a well-timed act of mid-scapular stabilization. Going forward, we would like to study and compare the timing of each muscular action and consider their implications.