Notes on S.I

Notes on S.I.


In the practical work, I often had the impression that the shorter leg was associated with anterior pelvic torsion and knee conflict. I became curious to investigate what seemed to potentially turn out as a meaningful pattern. Curiosity was joined by a need when sometimes unexpectedly the legs shifted their relationship of shorter to longer around or when the difference actually became worse. Certainly, resorting to an attitude where ?straightening out? the body would take care of the problem is unsatisfactory. And indeed, it sometimes did not do that. The inquiry presented here, examining structure from the parameter of the leg difference in length, does not solve the problem, but it may shed some light on the issue and lead to more poignant questions, the treatment of which may aid in becoming more specific technically.

The term ?shorter leg? is used to describe the statistical normality that everyone has a shorter and a longer leg. The fact doesn’t need further proof, as it derives logically from the triviality that there exists no absolute symmetry in biology. The concept belongs therefore to the physiological range of ?normal?, as opposed to the ?short leg?, which would mean a difference in leg length which is medically relevant – has acquired ?disease value? – by virtue of producing subjective or objective symptoms or threatening to do so in time. This is quite a different and tricky problem because at least in severe cases, the body has nearly exhausted its capacity for compensation, and dabbling around in it always threatens to unbalance a subtle system of equipoise.

The objective of comparing the length of the legs necessitates a definition of what constitutes the ?leg?. While it is commonly agreed that the leg ends at the ground below, it is less easy to define where it starts on top. In a purely skeletal view, the head of the femur would be taken. But functionally, the psoas takes it up to L1, and structurally it is of course a banality to state that the leg starts at the top of the head. In orthopedics, leg length is clinically measured by comparing the iliac crests. Boards are then put under the foot on the lower side, and the thickness of the boards necessary to bring the crests on a horizontal is considered the difference in leg length. Discrimination is very sharp with this method, and more ?sophisticated? methods, using X-rays or instruments, have not been able to improve on it. It can be held that it is an indirect measurement – or that the pelvis forms part of the legs -, but it is extremely sensible functionally. For, it takes into account that the difference in leg length is mostly relevant for the strain it induces in the upper body. And as the ?backbone? of this is the spine, the base of which is the sacrum, equalizing the iliac crests is of much more practical value than discussing where the leg ends on top.

To the naive mind, the simple question arises soon of how the pelvis adapts to the heads of the femora which are never exactly on a horizontal. For, it is evident that the acetabula have to fit them like caps. Four possibilities come to mind:

1. Rotation of the pelvis as a whole will lower one acetabulum if there is a marked anterior tilt. With a horizontal pelvis, the axis is vertical, and rotation around it will rotate every part in horizontal planes and so would not help any.

2. Sideway tilting of the pelvis would help, but it would take the sacrum along and impinge heavily on the spinal configuration.

3. Vertical shearing at the iliosacral joints would disrupt the integrity of the pelvis unduly.

4. Torsion.

Torsion seems to form the physiological adaptation mechanism to different heights of the femoral heads. The argument may sound somewhat teleological, but in fact torsion inflicts the least damage to the structural integrity of the body. As its axis is roughly horizontal and through the centers of the iliosacral joints, this puts it behind and above the axis through the acetabula. So, a relatively distinct difference in heights of the acetabula will be accompanied by a minor twisting of the sacrum. And especially the so important horizontality of the two lumbosacral articulations will be disturbed minimally. The argument is supported by the experience that torsion is always present, but that when its capacity for adaptation has been exhausted, the other mechanisms come into play. And that seems to be exactly the point where pathology sets in.

The more general question arises of where and how the shorter leg loses length relative to its longer brother. It is convenient to consider the problem under the premise that the bones of the legs are symmetrically identical. This is not to say that the shape of the bones is a quantity negligeable but only that it expresses the major concern for the role of the fascial web as a structural factor from a practical point of view. Structure is then meant in a narrow sense as the geometry of the connective tissue network. It may seem a little ironical that this relegation of the skeletal factor to a secondary cause is essential for understanding better the fascial geometry – and for this, the idealized bones form the primary indicators.

Structurally, the leg consists of three compartments or blocks: the foot, the lower leg, and the thigh. In a fascial context, it is evident that the foot is the most important as its configuration and height are determined much more by the connective tissue than by the simple size of its component bones. For the lower leg and the thigh, a difference in the length of tibia and femur has much more immediate consequences for the length of the compartment as they run close to its axis. But it should be noted that the two bones do not coincide exactly with the axis of the limb. This is evident for the femur, where the head and the neck go strongly lateral, the shaft medial. But also the tibia is not exactly in line, as its proximal center may be in the center of the knee joint, but its distal center is medial to the center of the ankle joint where also the fibula takes part. Wolf Wagner gives some good reasons why bone axes should not be in line with the axes of the limbs(1).

Structurally speaking, with the premise of the identical bones in mind, a shortening of the compartment can only be due to a deviation of tibia or femur from normal. The normal axis of the leg – not of the bones – is defined by a vertical line through the centers of the hip, knee, and ankle joints. Deviation of tibia and femur will reduce the length of their respective projections on the vertical and will therefore be effective in shortening the compartments in a structural sense. An exception must be mentioned here where the bones go exactly in line with the vertical, which would obviously maximize their projection on the vertical. This is not of practical importance, however, because it doesn’t seem to be part of compensating mechanisms, and when they do align so accidentally – always with a lot of rigidity -, the situation is soon changed by unspecific work as the bones start to slide in the direction of normal.

A Skeletal Hypothesis of Lower Girdle Structure

Besides the first premise of symmetrically identical bones, the hypothesis rests on the additional premise that the bones are absolutely centered in their joints. It is admitted that this is probably never the case, even if the deviations usually do not attain the status of subluxation. But it should also be noted that it would represent an ideal state of matters in accordance with the basic economic premise of Rolfing. However, Rolfing is not about harmonizing given particularities, intrinsic to the body, but about organizing it with respect to the framework given by gravity. On the other side, a structure ?straightened out? perfectly will not be stable if joints are torqued too much in the process, and the short-term goal is always a compromise somewhere between organizing the body maximally with respect to the verticals and the horizontals and respecting some harmony necessary within skeletal restrictions.

The hypothesis takes the torsion of the pelvis as given and its starting point. The description given earlier must be modified, however. Torsion is not a matter of inverse rotation of two elements around a common axis. It is a three-part affair, involving the ilia and the sacrum. To simplify matters, the twist of the sacrum is neglected – although it actually favors the objective of lowering one acetabulum and raising the other. Then, the perpendicular line through the center of the iliosacral joint, around which both ilio-ischio-pubic complexes rotate, is from lateral, dorsal, and cranial to medial, ventral, and caudal for both sides. For the three clinically relevant points, the movement is as follows:

– the anterior superior iliac spine – with anterior torsion -, goes forward, down and lateral; on the side of posterior torsion, it goes back, up and medial,

– the posterior iliac spine – with anterior torsion -, goes forward, up and lateral; on the posterior torsion side, it goes back, down and medial,

– the tuberositas ossis ischii – with anterior torsion -, goes back, up and medial; on the posterior torsion side, it goes forward, down and lateral.

The situation must not be confused with the tilt of the pelvis as a whole, which is around a truly horizontal axis, and which is usually accompanied by internal rotation for the anterior tilt, external rotation for the posterior tilt.

A simple indicator of pelvic torsion is the inguinal ligament. As it is a structure within the ilio-ischio-pubic complex, it follows its spatial movement faithfully. On the side of anterior torsion, it is more horizontal and wider, on the side of posterior torsion, it is steeper and narrower. The acetabula also follow the direction of torsion, and their axes

– go down and forward on the side of anterior torsion,
– back and up on the side of posterior torsion.

This predicts for the femora that

– with medially rotated femora, the stronger rotation is on the side of anterior pelvic torsion,
– with laterally rotated femora, the stronger rotation is on the posterior side of pelvic torsion,
– with mixed rotation, the medial rotation is on the anterior side of pelvic torsion.

<img src=’’>

Rotation of the femur is around the ideally vertical axis through the tip of its head – the intersection of the axis of the head with its circumference – and the center of its articular surface at the knee. Rotation would then not have any noticeable effect on the length of this segment. Deviation in the sagittal plane – flexion – would however shorten the projection of the bone on the vertical and with it the segment. But it only seems to play a role under pathological conditions, where the longer leg compensates by bending at the knee, in this way giving up some of its unbalancing length at both the femur and tibia levels. In the frontal plane, where deviation would also be effective, the femora are usually symmetrical enough to be neglected. But it should be noted that in bow legs, the abduction of the thighs brings the shafts of the femora more toward vertical and will actually make this compartment a little longer, while adduction increases the obliqueness of the femoral shafts and will shorten. This may be a factor in rare cases with a combination of a markedly abducted and an adducted thigh. An unforgettable example of this configuration – which still seems to haunt those who had to defend against him – is Garrincha, the right wing of the Brazilian world champion football team.

At the level of the lower leg, deviation in the frontal plane – especially ?abductions? – is often more prominent than rotation and could be relevant for the length of the segment. For the foot, as stated already, the height is determined almost exclusively by the connective tissue.

On rather flimsy evidence, and without being able to prove it, an extrapolation of the hypothesis can be construed. In a random body, where the shape of the legs is usually fairly discordant, the feet could be the primary substrate for compensating the difference in leg length, – if the point is taken that this would make more sense than adjusting to different heights of arches by an awkward twisting of the legs. But in integrated bodies, where the legs approach normal and are more equalized, the height of the feet would be more directly related to pelvic torsion, whether causing it or compensating for it.


22 clients were photographed. For all except one a checklist was filled out. Because of sometimes questionable discrimination of the parameters determined and my lacking experience with it, findings can only be presented tentatively. The pictograms which accompany the photographs shown have four horizontals for the hip, the knee, and the ankle joints, and for the ground. The pelvis is symbolized by an angle with the steep arm for the posterior torsion side, the flat arm for the anterior torsion side. The femora end above or below the knee horizontal according to the difference in height above ground. A medial hook stands for internal rotation, a lateral one for external rotation. Two hooks are used to designate the side with the stronger rotation. A hook at the tibia also stands for rotation in such a way that femoral and tibial hooks going in opposite directions denote a knee conflict. The relative height of the ankles is given.


The anterior torsion side shows a wider and flatter inguinal ligament than the posterior side. The iliac crest is usually lower and farther out on the anterior torsion side, but sometimes the crests are higher than the picture would suggest. Pelvic rotation, which is usually in the direction opposite to the one of the torsion, may make the posterior torsion side look wider. The inguinal area just above the ligament is wider and shorter on the anterior torsion side, narrower and longer on the posterior side. The upper line of the underwear may help, but it can also be deceiving when it goes over the iliac spine on the lower side but doesn’t make it on the higher side.


Rotation was determined by comparing medial and lateral epicondyles. Flesh bulges medially are sometimes deceptive. The patellae go sometimes in the opposite direction, taking a short-cut with the rectus. The proximal aspect of the adductors is often the most reliable sign: back and deep with internal, out and forward with external rotation with the adductors showing broadly. But a conflicting pelvic tilt may obscure the picture. The relative anteriority of the soft tissue below the pubes is a reliable sign for the relative degree of rotation.

Lower Leg:

Only the more strongly conflicted tibia is indicated when both sides are in conflict. Femoral and tibial hooks going to the same side should reduce overall height, but they don’t designate a conflict in the strict sense. They might rather be called ?hypercongruent?, giving an added momentum at the knee in the same direction as the femur rotates.


4 of the 22 clients show a left anterior pelvic torsion, the others are right anterior. An attempt at relating this asymmetry to the asymmetry of handedness was discontinued when two lefthanders among the first 14 clients showed the usual right anterior torsion. An earlier impression that such a relationship does not exist was so confirmed. The reason for the asymmetry is not known.

Homogenous shorter leg:

Only 6 clients show a homogenous shorter leg consisting of anterior pelvic torsion, a lower knee joint, and a lower ankle joint (Photos 1-3). 5 of them also have a knee conflict in the shorter leg, with the one exception shown by Photo 3. The count does not support the hypothesis, especially when one considers that the examples shown demonstrate a rather marked left-right imbalance.

<img src=’’>

<img src=’’>

<img src=’’>

Reversal at the hip level:

7 clients show an anteriorly torqued ilium higher than the posterior torsion side (Photos 4-6). This seems like a rather weak support for the hypothesis that the anterior torsion side should be on the side of the lower leg (15 clients). However, it seems remarkable that 6 of them show a complete reversal at the hip level in that their knee and ankle joints are also higher. Photo 6 is the exception.

<img src=’’>

<img src=’’>

<img src=’’>

Reversal at the knee level:

6 clients show this configuration. With 3 of them, the reversal is complete, i.e. the ankle is higher where the knee joint is higher (Photo 7). The other 3 are ?incomplete?, i.e. the ankle is higher on the side where the knee joint is lower (Photo 8).

<img src=’’>

<img src=’’>

Reversal at the foot level:

Only 3 clients show this inversion where the ankle is higher while the knee and the hip are lower and the anterior torsion is on the side of the shorter leg (Photo 9).

<img src=’’>

Rotational pattern:

It seems amazing that 18 of the clients conform with the prediction concerning the relationship of the rotations of the femora. Photo 3 shows one of the 4 exceptions. This could be taken to support the practical experience that front-to-back balance, of which rotation is an expression, has primacy over left-to-right issues ‘.

Photo 10. shows an irregular hip reversal with left anterior torsion.

<img src=’’>


It should be noted that the inquiry presented here does not allow to answer the question of which segment it is where the shorter leg loses length. But it does seem to support strongly the tenet that front-to-back balance – as expressed by rotations in the front view – is of much greater importance than side-to-side issues.

The findings raise a question again which was implied in the hypotheses. It concerns the notions of compensations. It seems reasonable to assume that very slight imbalances find their way into the structure and go ?unnoticed? by the ?deep? structural level. This would mean that the body simply behaves, i.e. moves, a little different without being upset structurally and adapts concomitantly. It would succumb to these slight alterations without reacting. Only when an imbalance occurs which is too much of a disturbance, the body would adapt with compensations. The general aspect of those clients with a homogenously shorter leg suggests that they might have a high tolerance for aberrations, going with them a long way. But from the present inquiry it cannot be excluded that they do adapt – in their upper body. If this were true, the question would arise as to which persons adapt to imbalances with compensations within the segments afflicted as opposed to those who use the rest of their bodies to compensate.

<img src=’’>

Reconsidering the premises taken, it is time to introduce the bones as a constituting factor again. For the reversal at the level of the pelvis, figure 2 shows a possible skeletal factor. This will not take care of all the phenomena, but it could be a determinant in some structures when considering that bones are very plastic in early life. Figure 3 shows a knee reversal being possibly dependent on early bone formation. This reasoning leads to a tentative structural hierarchy as shown by Chart 1. The relationship would be such a one that the intrasegmental imbalances, more isolated and less dependent on overall structure – possibly established early – would only be accessible after some ordering of the blocks. On the other hand, stacking the blocks would be limited by irregularities of the blocks themselves. This could explain the phenomenon that some intersegmental order and lengthening often happens fast initially, whereafter the process starts stalling and the structure resulting from a session actually falls back on where one started. Then, intrasegmental ordering would be indicated: the ?squaring? of the blocks internally.

<img src=’’>

Of course, this switching over from whole-body integration to working on local causes and vice versa is part of everyone’s practice. The appearance of ?deeper? aberrations, when pelvis torsion becomes strongerg., would be the sign to change over to working on the ?intrinsic of blocks. For, the advancement of overall structure can only be supported to the extent that the segments themselves are ordered.

<img src=’’>

The relationship between intrasegmental configuration and the shape of the bones seems similar to the one between intersegmental and intrasegmental order. Intrasegmental organization may be very advanced, but it will not be stable if it is too much ahead of the conditions given by skeletal shape. As bones change shape very slowly, it appears that there is nothing to be done about them – granted that the connective tissue is organized optimally – but to reinforce the electric phenomena that seem causal in altering the shape of bones by instructing appropriate movement – or by surgery.


1. See his paper in this issue of the  Notes on SI

To have full access to the content of this article you need to be registered on the site. Sign up or Register. 

Log In