The Theory of Lower Spinal Rotation:
How It Serves As A Style-Independent Description of Waist Powerby Tim Niiler Ph.D. and Henry P. Gong
Descriptions of spinal power generation in the martial arts have a tendency to be style specific, making comparison between styles difficult. We have studied the rotational kinematics involved in training spinal power generation in Shaolin Longfist Kung Fu and present the results using the standard language of exercise science. It was observed that proper generation of power required excellent alignment of the spine and relied on the muscles of the abdomen, and that this is not inconsistent with other martial styles.
Describing the Generation of Power
The generation of power in martial arts is a compelling topic, largely because of the ambiguities used in describing the technique in detail. Some styles claim to use the waist to generate power, while others describe using the pelvis, hips or spine. While some of these descriptions may truly be style-dependent, nearly all of them suffer from a lack of anatomical references which specify in detail precisely which body parts are moving in coordination and which muscles control them.
There is, perhaps, some justification for generalizing when teaching the public as few students of the martial arts are likely to have had anatomical training. However, these very general descriptions of power generation often appear even in detailed articles by professional martial arts instructors. This raises the question of whether the descriptions given by such writers are stylistic nods to tradition - implying no further detail exists, or whether the descriptions are simply limited by the author’s knowledge.
In most general popular descriptions, practitioners are told that the generation of power is dependent on a specific part of the body. Instructors using this method will tell the student to use the hips or use the waist, but do not tell them how.
An example of this may be found in /Bruce Lee’s Fighting Method/. In describing power generation, he states: “Your whole body should participate in the impetus - your hips, shoulders, feet and arms… The punch originates not from the shoulder but the center of the body” (Lee & Uyehara, 1977:68). However, this description only hints at the timing of the movement without describing the actual mechanics. Rotation is not mentioned.
The description given by the International Wu Shu Federation involving the “three external coordinations” detailing the coordination in movement between the shoulders and hips, elbows and knees, and hands and feet, is slightly better (International Wushu 1995:6). Yet, considering that its audience consists largely of professional wushu judges and coaches, many of whom have degrees in physical education, the description offered in this manual is decidedly vague.
Articles that describe both the timing of power generation and the part of the body used are on a slightly better level. One excellent example of this is the article by J. Chaudhuri (1995) who describes power generation in the Wing Chun form Sil Lim Tao. He describes power as being initiated at the knees to control the hips and subsequently the shoulders. In addition he describes the spinal column as the “motherline” about which all motion occurs. Yet, in spite of a thorough description, he ultimately relies on terminology specific to Wing Chun. When comparing power generation in Wing Chun to that of Chen style taijiquan, Chaudhuri’s terminology changes from what he had previously defined to something more reminiscent of the taiji classics: “A major source of the great moving energy in the Chen style comes from uncoiling, spiraling-upward movements with the waist playing a major role” (Chaudhuri, 1995: 75). Unfortunately, this classical description is vague in the extreme.
When writing descriptions, in addition to specifying the parts of the body involved in the motion and the timing, it is important to use standard language accessible to people anywhere. We suggest using the commonly accepted terminology of anatomy which is standard across all languages. While it may be argued that language based on exercise science (i.e. Latin) also excludes outsiders, you can find Latin based anatomy books in almost any bookstore anywhere in the world. This is not the case if you wish to acquire materials to study the basis for a specific esoteric style.
Examples of the use of this language in describing power generation do exist and provide unambiguous and universal descriptions of what is happening in the execution of a given technique. Hobusch and McClellan (1990) offer an excellent description of the sequence of movement and body parts involved in the execution of a karate style roundhouse kick. While their description is certainly too terse for the average practitioner, it is a perfect template for instructors who can then amplify it as needed to help students correct their kicking technique.Van Gheluwe and Van Schandevijl applied this terminology when describing their analysis of the reverse punch (Van Gheluwe & Van Schandevijl, 1983). Pieter’s review of their study summarized their findings in a clear and concise statement: “They found a lateral and rotational deformation of the spine that started at the hips and moved up to the shoulders” (1994: 29).
An Anatomical Perspective: Hip Power vs Lower Spinal Rotation (LSR) ®
Hip Power, as it is usually understood is generated by the legs which are rooted to the ground. Regardless of the fighting style, hip power requires forward rotation of the pelvis to occur in the coronal plane (the plane intersecting the waist, parallel to the floor). This rotation is not due to the abdomen as commonly believed. Instead it is the result of the rooted legs exerting a force on the pelvis. In this motion, the adductors of the forward leg pull against the pelvis, causing it to rotate. This may be supported by a pushing motion by the back leg (knee extension) which causes a pivoting motion about the forward hip socket. Finally, the gluteals of the rear leg (including the /obturator internus/ and /piriformis/) may contract to help square up the hips with respect to the back leg (Figure 1). This pivoting motion may in some cases be as great as 90 degrees.
In comparison to the motion created by hip rotation in martial arts, the motion in the lower spine is quite small. This is because of the limited range of motion of the lumbar spine. Each vertebrae has a full range of rotational motion between 2.5 to 5 degrees per vertebrae. Summed along the 5 lumbar vertebrae, the total average range of motion, accounting for variation between the specific vertebrae, can vary between 13 to 17 degrees (Nordin & Frankel, 1989). As it is used in martial arts, rotation of the lower spine is not a substitute for hip power; rather, it supplements that power, linking the upper and lower body. This usage of the lower spine to link the torso with the legs has been designated Lower Spinal Rotation (LSR®) by Henry Gong. The phrase is anatomically preferable to “waist power” since it describes both the specific motion and the location where the motion originates. Likewise, it is and adequate description of the torso portion of power generation given in several of the articles mentioned above, including Chaudhuri (1995), Hobusch and McClellan (1990), and Van Gheluwe and Van Schandevijl (1983).
In theory, rotation about the lower spine may be caused by contractions in several different muscle groups. The most predominant muscle groups involved are the abdominals; more specifically, the external and internal obliques. These muscles attach at the ribs and fascia at the front of the abdominal wall. You will use the left obliques to turn the torso to the left and the right obliques to turn the torso to the right. Supporting muscles used in lower spinal rotation include the /transversospinalis/ in the back which nestle very close to the spine. While involved in LSR, it should be noted that these muscles are not nearly as strong or massive as the obliques, and therefore can only be considered to have a supporting role. A final group of muscles, the /erector spinae/, should be mentioned as they are considered to be the major postural muscles that maintain a straight back. Without some activity from the erector spinae, the back would collapse upon action from the abdominals.
Quantifying Lower Spinal Rotation
Although when presented anatomically, the execution of LSR seems fairly self evident, it is still valuable to directly observe these motions and quantify them. Such observation and analysis helps us to understand the motion’s timing and the reliability with which it can be repeated.
In modern biomechanical technique, three types of measurements are typically made: kinematics, kinetics, and electromyographs (EMGs). Kinematics include body positioning, velocities and relative joint angles as they vary with time. Kinematics are measured by simultaneously filming an action from several angles using multiple high speed cameras. Kinetics are the forces acting on the body and their variation with time. They are usually measured using force plates embedded in the ground and forces at the joint centers are inferred from calculations when possible. Finally, EMGs are measurements of muscle activity as it varies with time. EMG data are collected using fine wire electrodes which are injected into the muscle or surface electrodes which stick to the skin are easily removable. While the less invasive surface electrodes are more commonly used, fine wire electrodes are more versatile as they can be used to probe deeper muscles.
In order to better quantify the kinematics and muscle activity involved in LSR, we filmed simple rotational motion using modern motion analysis techniques, with Henry Gong executing rotation about the spine in a stable stance. This motion involved a minor turn followed by a more pronounced turn, first to the right and then to the left side. Mr Gong was marked for kinematics using the 21 marker Helen Hayes ™ marker set (Figure 2).
The kinematics were obtained using a Motion Analysis Corporation™ six camera 3D HiRes™ system. Since the motion was not excessively fast, film speed was set at 60 frames per second. The EMGs were obtained using four channels of a Transkinetics TXRE-206™ EMG amplifier connected to a computer via a data acquisition board. Since surface electrodes were used, we chose to examine the lumbar area of the erector spinae and the external obliques . The internal obliques were inaccessible to surface electrodes, so were not considered in this experiment. For the erector spinae, electrodes were placed at about L3-L4 (3rd and 4th lumbar vertebrae) and within two inches of the spine. The spinous processes were used as the sites for the ground electrodes. For the external obliques, the sensory electrode was placed about a hand’s width from the rectus abdominus at the level of the umbilicus. Data were collected for four trials.
Data were analyzed using Orthotrak Clinical Management™ software. The kinematic variables of interest were the trunk’s coronal plane rotation its forward tilt, and its lateral tilt (Figure 3). Also of interest were the EMGs of the erector spinae and external obliques as output by Orthotrak. EMG detection sensitivity was set at 50% maximum input in order to more clearly distinguish when the muscles were active.
The correlation of muscle activity with the rotational motion of the trunk is of primary interest. A graph overlaying rotational motion with the EMG signals from the external obliques illustrates the average motion and muscle activity present in this rotary motion (Figure 4). Ideally, muscle activity precedes visible motion and fades shortly before the completion of the activity. This was most evident in the initial turn to the right in which there was a consistent 1/12th of a second lead time in muscle activation. While this pattern was seen later in the movement, it is not present for every turn. In some cases the muscle activity lagged the movement. This implies an alternate mechanism for the return to center. It is possible that some of this motion was due to the elasticity of the muscles returning to the trunk to a more neutral position. But it is just as likely that other muscles were involved, and their activity was undetected by the surface electrodes. These muscles could include the internal obliques which are deep, relative to the external obliques, and the iliopsoas which connects the lumbar spine to the femurs.
Examination of erector spinae activity in one representative trial containing nearly two complete turning cycles (Figure 5) showed that right erector spinae activity accompanied rightward tunring 43% of the time, and left erector spinae activity accompanied leftward turning 42% of the time. However, right erector spinae activity accompanied leftward turning 22% of the time, and left erector spinae activity accompanied rightward turning 59% of the time. Simultaneous contraction of both right and left spinae occurred 18% of the time.
When using 20% instead of 50% of maximum as the minimum activity detection level for the EMGs, it became clear that both right and left side erector spinae are almost always active, even if at low intensity. These figures indicate no correlation between erector spinae activity and the direction of trunk rotation, but are consistent with the usual role of these muscles as stabilizers of the back.
As stabilizers, they were obviously effective. Average forward tilt of the trunk with respect to the pelvis was very consistent across all trials (Table 1). Standard deviations from these values were quite small, ranging from 1.0 to 2.6 degrees. For lateral tilt of the trunk with respect to the pelvis, the data showed a slight (0-3 degree) lean to the right, but again the standard deviation about this value was small, ranging from 2.2 to 4.0 degrees (Table 2).
Trunk rotation is the basis for many martial arts movements including striking, pulling and throwing techniques. The collected data indicate that rotary motion is initiated by muscles in the abdomen, and that the muscles of the back play a supporting role in holding the spine erect. This is significant from a training perspective. While most martial arts teachers emphasize lower abdominal conditioning via crunches and other related exercises, few work to train the rotary muscles of the trunk in the postures that are utilized in form or fighting. Adding such training will likely prove to be as valuable to martial artists as to the athletes in many other sports that utilize specific muscular strength training as an essential element.
The timing of techniques is a crucial component for this training. An experienced martial artist such as Mr Gong was capable of initiating a change in direction from the lower spine at least 3.6 times per second as shown by data (Figure 5), although personal visual observation has shown him to be capable of movements much faster. This ability to change direction at the core of the body relies on rapid initiation of techniques and a short gap between muscle activation and actual rotation of the body. Longer activation times result in excessive tension in the muscles which drains needed energy, reducing available power.
Muscles cannot push. They can only pull. Muscle contractions cause bones to swing. As such, torque and not power is the defining quantity in the discussion of martial “power”. Torque is generated by applying force to a level. The distance between the place the force is applied and where the lever pivots is called the moment arm between the pivot point and the external load or resistance (usually provided by the opponent). Although one can increase the muscle force by conditioning the proper muscles, this is an exercise in diminishing returns. But decreasing the moment arm between the external load and the pivot point is something that can be controlled via posture.
As demonstrated by the data, LSR therefore encourages upright posture so that the spine becomes the pivot point in the body’s motion. In this mode, the body acts as a level (Figure 6). Deviations in the alignment of the lower spine can therefore have the effect of lengthening the moment arm of the external load and increasing the net torque exerted by it.
Procedure and Results
Although a skilled anatomist can often observe a movement and describe exactly how it happens, in some cases, more precise observation is needed. Muscles thought to be uninvolved could turn out to be important in the completion of a movement while those thought to be prime mover may only play a supporting role. The reason for this is what is known as co-contraction. When several muscles play a role in a motion, their activation pattern may actually be quite complicated. It is much like when you go to the store, you have a number of good ways to get there, depending on what is important at the moment. Just having a street map will not enable someone else to predict how your will get there at any given time. In much the same way, the anatomist may not be able to predict precisely which combination of muscles are utilized for a given movement without direct observation.
This means you have to observe the joint angles and muscle activities involved in the motion. By determining that a muscle activation leads to a related change in joint angles, it is then possible to say that the muscle activity caused the movement. However, sometimes when muscles co-contract, it may be that the muscles assist the movement, only by stabilization the involved body part, and are not directly involved in the movement itself. This situation is further complicated by the fact that muscle activation should ideally preceed motion since activation occurs when the electrical contraction signal from the brain arrives at the muscle.
Our observations showed that the erector spinae muscles were contracting almost constantly regardless of the direction of motion. This seemed to indicate that these muscles were more useful for spinal stabilization but were not much used for the initiation of the movement. The stability of the spine was not in doubt as variation in vertical alignment did not exceed a maximum of 4 degrees.
We also found that although some rotational motion could be explained by the activation patterns of the external obliques which lie in the front of the abdomen, some remained unexplained. This led us to believe that deeper muscles without sensors, such as the ilio-psoas, were partials responsible for lower spinal rotation. If the ilio-psoas are indeed involved in LSR, this adds weight to the idea that LSR couples the upper and lower body as the ilio-psoas attach to the spine and the femurs.
This study has provided some key details in the execution of Lower Spinal Rotation including timing information, muscle activation and kinematics of the basic turn. Although coordination patterns of the movement were described, fine wire electrodes were not utilized and so, detailed information on muscles other than the erector spinae and external obliques could not be obtained. The study only focused on the basic rotational motion and not specific martial techniques. A study of basic punching, or pull techniques as used in grappling, could provide details of the coordination patterns necessary for optimal performance of these techniques. While the focus of the study was on Lower Spinal Rotation, we acknowledge that hip motion is also important in power delivery, and that LSR should be used to link the upper and lower body to increase available power.
Chaudhuri, J (1995). Defending the motherline: Wing Chun’s Sil Lim Tao. Journal of Asian Martial Arts, 4(4), 70-83.
Hobusch, F. & McClellan, T. (1990). The karate roundhouse kick.
National Strength and Conditioning Association Journal, 12(6), 6-9, 84-88
International Wushu Federation (1995). Wushu International judge teaching material.
(H.X. Huan & X.Z. Luo, Trans.). Beijing: International Wushu Federation.
Lee, B. & Uyehara, M. (1977). Bruce Lee’s fighting method: Basic training.
Santa Clarita, CA: Ohara Publications.
Nordin, M. & Frankel, V. (1989). Basic biomechanics of the musculoskeletal system.
2nd Edition. Philadelphia: Lea & Febiger.
Pieter, W. (1994). Research in martial sports: A review.
Journal of Asian Martial Arts, 3(2), 10-47.
Van Gheluwe, B. & Van Schandevijl, H. (1983). A kinematic study of the trunk rotatin during a gyaku-zuki using tilted-plane cinematography. In H. Matsui & K. Kobayashi (Eds),
Biomechanics VIII-B. International Series on Biomechanics, Vol. 4B. Champaign, IL: Human Kinetics Publishers.
first published in:
Journal of Asian Martial Arts
Vol 14, No 1, 2005