I invented this game to help my students learn about the movements at each joint, viewpill and how to use anatomical language to describe movement. I eventually want to make an interactive game for the blog, but thought some of you might enjoy this as it is. As a teacher, I typically will be the robot first, and the students each take turns programming. I start kneeling, with my left index finger touching the ground , and my right hand shaped like a cup on the right thighas the start, and my left index finger in the cup as the finish. Then I have my students work in pairs. I troubleshoot any problems or errors my students make, and answer questions as they are working through. Let me know if you have any suggestions about how to make it better, or if you come up with other variations.
You can use this PDF (Joints Categorized by movement) as a guide to what movements are possible at each joint. The GetBodySmart website has excellent resources to help you learn more about these movements, but they focus on the level of muscles, rather than the joint. I hope to eventually provide a more simple explanation.
Get the robot from the start position to the finish position
Divide into pairs. One person is the ‘programmer’, the other is the ‘robot’. You can also have more than one programmer, and during ‘programming’ the programmers will take turns.
The robot demonstrates a starting posture and announces “This is the start”, then demonstrates a finishing posture and announces “This is the finish”.
The programmer then tells the robot to move, one joint at a time, ONLY using anatomical language, for example, “Flex your left elbow”.
The robot will then move the joint in the appropriate manner, i.e., flex the left elbow slowly.
When the position for that joint is satisfactory, the programmer says ‘Stop!’ If the robot moved too far, the programmer(s) must correct the over-movement on the following turn.
If the programmer uses incorrect language, for example asks the robot to ‘Flex your arm’ (not a joint), or to ‘Abduct your elbow’ (not a possibility at that joint), then the robot must tell the programmer ‘I’m afraid I can’t do that. Dave.’ and the programmer must revise the command.
The programmer continues until the robot reaches the finish position.
Switch roles: the programmer becomes the robot, the robot becomes the programmer.
The programmer decides the start position for the robot, and does not tell the robot the end position. Surprise!
 It is very important to choose postures that can be held for a long time, since programming can be challenging. It is equally important to not choose postures that require shifts of center of gravity, since these moves are very difficult to execute one joint at a time: your normally make many changes at once when shifting weight.
 Invented at a retreat by some of my students from Vermont.
So, I have been running a bit more than usual recently, kilometers creeping up to around 60 per week, and long runs of 20 each Saturday. I have also been stretching a little less, you know, the perils of everyday life. I began feeling a dull, constant ache in my buttocks, which then started creeping down my leg. I recognized this in myself as being piriformis syndrome. All of this got me thinking about the piriformis muscle, and its relationship to the sciatic nerve, and I thought it would make a good topic for a post.
The piriformis (piriform = pear-shaped) muscle is typically considered one of the 6 deep outward rotators of the hip, along with the obturator internus and externus, gemellus inferior and superior and the quadratus femoris. The piriformis, obturator internus and gemilli share a more or less common tendon on the greater trochanter, common innervation, and seem to work together under many circumstances, leading some researchers to call them the quadriceps coxae. (Standring et al, 2008). Of course, the big power rotator is gluteus maximus, but the deep rotators, like most deep muscles, are slow twitch or phasic muscles, and are therefore involved in most outward rotation.
When the femur is in anatomical neutral, the piriformis always acts as an external rotator. This is because the origin is on the anterior part of the sacrum, and the insertion is on the greater trochanter of the femur. The muscle runs posterior/medial/superior to anterior/lateral/inferior, and the angle of pull will draw the greater trochanter posteriorly relative to the femoral head, which results in outward or lateral rotation, and weak abduction. Fortunately for the sake of movement, but unfortunately for easy understanding, the femur can be flexed and extended, adducted and abducted, inwardly and outwardly rotated, allowing for many movement possibilities.
All any muscle really wants to do is get its two ends closer together. So when because of movement of the femur, the greater trochanter changes its relationship to the sacrum, the line of pull will change. In fact, at a certain point of hip flexion, piriformis switches from being an outward rotator to being an inward rotator. This is not just a minor weirdness or problem for biomechanics students, but has clinical significance, since many people wish to stretch the piriformis, and some of them suffer from a condition called piriformis syndrome. When the femur is moved to a certain point of flexion, the relationship of origin and insertion is changed: the direction becomes posterior/medial/inferior to anterior/lateral/superior. In this position the angle of pull will cause the greater trochanter to move more medially and superior, and combined with a fixed point of rotation in the socket, will result in inward rotation.
When I first started investigating this idea years ago, everybody seemed to be referencing The Physiology of Jointsby I.A. Kapandji (1970) who argues the change occurs at around 60° of flexion. I am a little surprised at the generous citing of Kapandji, a source I suspect it somewhat like the bible, where everyone quotes it, but few have actually read it. Kapandji is certainly a venerable name in the pantheon of biomechanical demigods, but even the revised edition of “The Physiology of the Joints: Annotated Diagrams of the Mechanics of the Human Joints” only has citations as recent as 1974. Some of the information was derived from cadavers at a time when technology and tissue baths may have not allowed for full range of movement, so that may have reduced observed angles, but even so, there is no indication Kapandji did any primary research: he appears to be citing someone else, unnamed.
Furthermore, Travell and Simons (who say the role of piriformis changes at maximum flexion, or ~ 150°) and Kapandji (~60°) cannot both be correct. If we average the difference, we come up with the modern idea of about 90-110° (Delp, et al. 1999, Dostal et al. 1986, and Neumann, 2010). These measures were obtained with modern radiographic, ultrasound and kinesiological techniques, and in my experience, are more consistent with in vivo clinical assessment and anecdotal evidence. (These also appeared in peer reviewed journals, rather than monographs.)
A paper by Pressel and Lengsfeld (1997) using a computer model of the human body that indicates the change from external rotator to internal rotator occurs at 70 degrees of hip flexion. Still, I think the angle of piriformis is more likely to make it an abductor at 70 than an inward rotator unless there is significant neutralization by the adductors (but that’s just a guess based on observation of movement and clinical practice). As Neumann points out, this is very easy to prove with a skeleton and a piece of string.
A really interesting paper by Vaarbakken and others was published in 2014. In the study, they used cadavers to look at the length of the piriformis muscle under combined conditions of flexion/extension, abduction/adduction and inward/outward rotation. By doing this they could also determine peak ‘moment arms’ for the muscle, a measure of when the muscle could be expected to produce the greatest force given its length and angle. They found two really important things. One, the piriformis is stretched the most when the femur is flexed to 105°, adducted by 10°, and outwardly rotated more than 25°. Two, the piriformis is actually disadvantaged in terms of power production when we are in anatomical neutral, but has its greatest moment arm as an extensor and abductor when the hip is flexed between 60-90°! In other words, friends, the piriformis’ main job is NOT as an outward rotator, but as a hip extensor and abductor when we are propelling ourselves from squats. Obviously this idea would need to confirmed with EMG studies, but still my mind is cautiously blown!
The important idea however is that what happens when the hip is flexed more than 90° determines what stretches the muscle. Normally, you just work out the opposite of a muscle’s function to determine its best stretch. So the gastrocnemius muscle plantar flexes the ankle and flexes the knee. Do the opposite and stretch the muscle: in the case of gastroc, dorsiflex and extend the knee. In the case of piriformis, if we only think of its job in anatomical neutral, then inward rotation and adduction will stretch it. But this is actually a poor stretch for the piriformis. A simple modification to improve this stretch is to slightly flex the knee. This position is similar to the F.A.I.R. (flexion, adduction, internal rotation) test for piriformis syndrome, which is hypothesized to actually trap the sciatic nerve in the notch (citation needed), and so this might not be the best stretch for the muscle, nor the safest test for piriformis syndrome.
Aside from being a biomechanical conundrum, the piriformis muscle also has a very important relationship with the sciatic nerve. The sciatic nerve is the largest single nerve in the body, made up of branches that emerge from the bottom two lumbar vertebrae and the front of the sacrum. The nerve is about the thickness of your thumb at this point, and it passes out through the sciatic notch or foramen, which is bounded above by the iliac bone and sacrum, and below by the piriformis muscle. Then, most typically the nerve will run under the piriformis and down the back of the thigh and leg, also being the longest nerve in the body. It receives sensory information about the skin, and provides motor commands to the outward rotators of the hip (but not piriformis or quadratus femoris), the hamstrings, and muscles of the lower leg and foot. At some point along its length, as early as immediately after emerging from the sacral plexus, and as late as the back of the knee, the sciatic nerve will divide into common fibular and tibialis branches.
This is one of the great “jazz” moments in the human body, where there is so much variation on a theme, you would think the forces of creation had just dreamt up Miles Davis and then started thinking about the piriformis/sciatic nerve relationship. In most instances, the nerve passes entirely behind the piriformis. In about 16% of cases, there is some kind of variation, with one both branches of the nerve passing in front, or through the muscle, or some combination one branch or the other passing in front, behind or through. (Roydon-Smoll, 2010).
If the sciatic nerve becomes damaged or compressed, it will cause intense pain, often seeming to originate in the buttocks and shooting down the back of the leg. Pain that is caused by the sciatic nerve in this way is called sciatica, and frequently occurs because of compression of the nerves as it emerges from between the vertebrae, because of a dysfunction of the skeleton.
A different possible cause of sciatica is overuse, overtightness or inflammation of the piriformis muscle, which because of its close relation to the sciatic nerve will cause it to become compressed, a condition known as piriformis syndrome. The syndrome, like all syndromes, is a collection of related symptoms that usually stem from a common cause, and which symptoms emerge will vary from person to person. The symptoms can vary a great deal, but are basically the same as the symptoms of sciatica: nerve pain in the buttocks which radiates down the leg, often with numbness and loss of function. It is always important to find the root cause of pain, and qualified professionals will use functional diagnostic tests or imaging to determine the cause.
Those who have sciatic nerve variations do not suffer from piriformis syndrome any more than other members of the population, however manual therapy may be more complicated for those individuals. Stretching remains beneficial, however piriformis stretches with inward rotation, flexion and adduction may actually apply pressure to the sciatic nerve because they necessarily bring the piriformis closer to the sciatic notch. Therefore stretches with the femur in flexion above 90°, with outward rotation and adduction may be preferable in the case of piriformis syndrome, and some good stretches can be found in yoga.
What body part a yoga pose helps varies greatly from body to body, and even from one side of the body to the other, but for the reasons listed above thread the needle (the outwardly rotated leg) and pigeon pose (the front leg) tend to help the piriformis greatly. Here is an article by Natasha Rizopoulos describing these poses. Her description of the poses is excellent, but she doesn’t really discuss the piriformis anatomy here. Another pose, the seated spinal twist may be too intense for some individuals with tight piriformis because it combines extremes in flexion, adduction and outward rotation. Modifications here to lessen the stretch could be very helpful.
Information from sciatica.org suggests that manual therapists should avoid direct pressure to the piriformis (or to any muscle they suspect of entrapping the nerve), since that will reflexively compress the sciatic nerve running underneath. Instead, by applying pressure to the inferior border of the piriformis and bowing it superiorly to the client’s ipsilateral (same-sided) shoulder, you can apply an effective stretch that may be safer for the nerve. Slow stretches of long duration will be less likely to cause inflammation or a reflexive contraction of the muscle.
Additional points here are that piriformis syndrome is often accompanied by sacro-iliac joint dysfunction, or tight iliotibial band, or sciatica of the spine, or tight hip flexors or all of the above, and it is important to understand these coincident problems. There is also biceps femoris syndrome, where one of the branches of the sciatic nerve is entrapped by the biceps femoris, but there is no reason why the piriformis or biceps femoris are the only possible causes of entrapment. In one instance, I had burning, ‘nerve-like’ pain running down the lateral side of my leg, just about where the common peroneal nerve passes under the distal attachment of the IT band. I applied a release technique to the IT band, and felt immediate relief, and I suspect it was yet another sub-species of sciatica. Again this is why it is important to connect with people who know more and to enlist their help.
For me, I have been applying a regimen of stretching, strengthening and self-massage, careful to apply appropriate pressure. I have gotten a great deal of relief, but things still aren’t quite right. I think I need some help with my sacro-iliac joint, and will be seeking care from the doctors and practitioners who will help me to move further along the continuum to full health.
Those of you who are interested in piriformis syndrome should look here for a really good discussion of the topic by people who have thought about it much more deeply than I. Sciatica.org is dedicated to the exploration, understanding, diagnosis and treatment of sciatica, and they provide many great resources.
So I was asked by a dance student about the different kinds of subtle movements in the knee joint. There is the obvious flexion and extension, but there are also a couple less obvious things. So what happens to the patella? Does it move with the tibia, or does it move with the femur? And what about that weird thing when I lock my knees? What is that all about? Both ideas are about relative motion: one part of the body is in motion relative to the other, and what we see as being in motion depends on our view point.
So first the patella during flexion. When we flex our knees, the patella maintains a constant relationship with the tibia: it does not move relative to the tibia. This is because the patelar ligament, when fully extended is of a fixed length. As our knee bends, the front face of the femoral condyles start in contact with the posterior surface of the patella, but as the flexion continues, then the distal ends of the femur have contact.
So you could say the femoral condyles glide under the patella, or the patella glides over the femoral condyles. It depends which end of the leg is fixed. If we sit in a chair and extend and flex our knees, the tibia and patella move together over the condyles of the femur. If we plant our foot on a step and step up, the end of the femur glides across the top of the tibia and back of the patella together. In this video, the patelar ligament extends to full length as it is placed under tension (I did not talk about this extension in class, because it is not important for our purposes). Then because the femur is fixed and the tibia is in action, the patella glides over the condyles of the femur. But relative to the tibia, the patella does not really move.
Additionally, the patella does a little side-to-side dance as it glides over the femur. It starts medial, and moves more laterally. This is because the posterior surface of the patella is forced the by tension on the quadruceps muscle and patelar tendon into the ‘intercondylar notch’ which is the indentation at the distal end of the femur. The patella moves laterally at the very beginning of the knee flexion.
Standing for a long time can be energy consuming, so your body likes to find biomechanically efficient solutions. One solution your body uses is to ‘lock’ your knees into full extension. In this way neither your quads nor your hamstrings really has to work to maintain stance. Your friends in grade 5 knew about the screw-home mechanism, and would run up behind you when you were waiting in line, put their knees just behind yours, and pop their knees into the back of your legs at the knee joints. Suddenly your knees were ‘unlocked’ and you would have to recover. Sadly, the teacher always looked just as you were getting revenge…
The screw-home mechanism works because in the last 20° of extension, the end of the femur glides relative to the tibia and moves into a position of relative stability. This rotation occurs because the lateral condyle of the tibia is smaller than the medial condyle, so the lateral side acts as a pivot point, and as the knee approaches full extension, medial side experiences a larger shift of the femur and tibia relative to one another.
What we say is rotating depends on which end is fixed, and whether we are moving INTO the locked position, or OUT OF the locked position.
As we are moving into full extension, that is ‘locking’ the knee into the screwed-home position, then the femur rotates inward if the tibia is fixed, for example when we coming to the top of a plié (that is bending the knees with the feet on the ground for you mortals). If the tibia is free, for example at the top of a jump on full leg extension, the tibia will rotate outward relative to the femur.
As we are moving out of full extension, that is ‘unlocking’ the knee from the screwed-home position, then the femur rotates outward if the tibia is fixed, such as when we begin our plie. If the tibia is free, for example if you were for some reason going directly from tendu dégagé devant to a retiré (for you non-dance types, standing on one leg, with your gesture leg fully extended in front with your toe barely off the ground,then moving to your toe at the side of the standing leg), then the tibia would rotate inward relative to the femur. This outward rotation of the femur/inward rotation of the tibia is cause by the popliteus muscle. In this video, the femur is fixed, so the tibia rotates laterally/outwardly as the knee reaches full extension.
There is one more small knee movement that I will mention briefly. It is the action of a small muscle called the ‘articularis genu’. Its job is to move the suprapatellar bursa out of the way so you don’t pinch it behind your patella when you extend your leg fully. The suprapatellar bursa prevents friction of the cojoined tendon of the four muscles of the quadruceps. And I am very glad for this action, because pinching a bursa can be very painful.
There are other subtle movements of the knee, since the knee is an enormously complicated joint, but these three are big enough that I can kind of ‘get’ their significance on a visceral level. The others will have to wait for another question, or a day when my mountain bike is not calling me!
To kick things off, I want to tackle a question posed by a friend on Facebook, and in so doing, I pray that I do not start a war with the venerable Katy Bowman, but maybe we can help each other learn more, and then go for waffles.
My friend read a blog on katysays.com, in which Katy says there is no such thing as the ‘iliopsoas’ muscle. My friend asked a bunch of us to say what we thought about that. At first I just reacted to the title of the article, and the evidence of that is in PART I of my response. But then I read a paragraph from the article, and several articles on the iliopsoas that made me want to say a little more in PART II. I provide links to the abstracts of these articles or the articles themselves. Unfortunately, unless you have an institutional library access, you might only get to read the abstract from the link, so although I apologize for that, I won’t apologize for using peer-reviewed sources only on this site. I encourage you to read any full article, including Katy’s, rather than take my word for it.
In Katy’s insistence that there is no iliopsoas muscle, she is being at little pedantic, but for me to say that is certainly the pot calling kettle black, so please, let’s get that on the table right away. Yes, of course there is no iliopsoas muscle, but in reality, the sun doesn’t ‘rise’ – the earth rotates. Still we say ‘iliopsoas’ and ‘sunrise’. Some of it is quaint convention, and some of it has to do with perception, and other parts have to do with the way things really are. I will leave sunrises for poets and cosmologists, and I will talk about muscles.
We have many muscles with special names because they share a common tendon. The quadriceps for example describes four muscles with a common tendon at the knee. The three vastii join to the femur and the rectus femoris crosses the hip. Only one of them flexes the hip, one of them pulls the kneecap laterally, one of them pulls the kneecap medially, but they all extend the knee. The pectoralis major is more radical, with sternal and clavicular divisions that can make the muscle act as its own antagonist in flexion and extension. There is also the triceps brachii, the triceps surae, biceps brachii, biceps femoris among others. Individual muscles with distinct and different functions, grouped together by a common tendon on one end.
In the case of the iliacus and psoas muscles, the iliacus inserts proximally on the upper part of the inside of the iliac bone, and the psoas major (don’t get started on the psoas minor, which is only present in 56% of the population, differs by race, side of body and sex…) has a proximal insertion on the 12th rib and the transverse processes of T12-L5 (ish). The two muscles co-join at the tendon which inserts onto the lesser trochanter of the femur, although some fibres of the iliacus continue further down the femur. In actual fact, most of the fibres of the iliacus insert not onto the lesser trochanter itself, but onto the tendon of the psoas. A clear case of riding in on another muscle’s coat-tails!
The fascia of the psoas muscle is continuous- and joins with- the fascia of the iliacus muscles as soon as it descends below the lip of the iliac crest. Psoas major and iliacus are both innervated by the second and third lumbar nerve, which I think is very important for our discussion here.
When acting together, the two muscles flex the hip, and may cause some adduction, and some inward or outward rotation of the femur, depending on the angle of the femur at the beginning of the movement. In terms of flexion (the principal action) they lift the femur if the pelvis is fixed, and cause one to sit up if the femur is fixed. Although EMG studies are difficult on these muscles since you need painful needle electrodes, and lots of money to compensate your volunteers, we do know that in hip flexion they always act together, and of course this makes total sense: any synergy of your body would not allow for a perfectly situated muscle to just ‘go along for the ride’. If it is there, and available, it will contribute. This is because they are innervated by the same nerves! The motor cortex makes a plan to move the leg, and uses all the possible muscles for that particular movement, it sends the command down the appropriate branch of the nerve, and the motor units are recruited, and voila! The leg moves.
I have seen no literature, either by pathology or by nerve block where the iliacus acts independently of the psoas, but it might be out there. On the other hand, the psoas does many things without the iliacus. The psoas stabilizes the pelvis in locomotion, can flex the lumbar spine if the thorax is in static flexion above T12, can hyperextend the lumbar spine if the thorax is allowed to extend or hyperextend above T12, can cause lateral flexion, and other complex movements. Its action is sometimes considered paradoxical or enigmatic, and its role overall is poorly understood given its apparent importance.
Are they separate muscles? Well, sure. But so are long and short heads of the biceps brachii. If Katy is going to apply this rule to the iliopsoas, she needs to apply it to all other multi-head muscles as well. And she should also stop saying ‘sunrise’.
I would also add that Katy doesn’t like calling the psoas a hip flexor because it does other things. That is not a just reason to make such a claim. A muscle can (and must) have than one function, otherwise you would need a bajillion muscles. As it is, a muscle’s function is contextually derived, and depends on what synergist and neutralizer muscles are available. Saying the psoas doesn’t flex the hip because it has other jobs is like saying the biceps doesn’t supinate the forearm because it flexes the elbow, or like saying Brian May isn’t a PhD in astrophysics because he was the guitarist for Queen. Well he is both. Silly rabbit, the psoas can do both.
Okay, I have now read Katy’s article more thoroughly, as well as a few research articles about the psoas, and I take back what I said earlier about her article: There are parts of it I dislike.
What I dislike in particular is the sentence that begins “But the psoas (unlike actual hip flexors) doesn’t tilt the pelvis anteriorly, it tilts the pelvis posteriorly relative to the ground.” There are a few problems here, not the least of which is her unclear nomenclature. When the ASIS moves forward relative to the pubic bone it is typically called anteversion. When the ASIS moves backward relative to the pubic bone, it is called retroversion. I believe she is saying the psoas does not antevert the pelvis, but it retroverts it.
We get into interesting territory now. Let’s go back to my earlier example of the pec major. The pec major has one common attachment on the humerus, just under the deltoid, but proximally, it has one attachment on the sternum and one on the clavicle. I could say: pec major is a horizontal flexor of the shoulder, and nothing else! Of course, the pec major (clavicular) is also a shoulder flexor, while pec major (sternal) is a shoulder extensor. I could stop there and say, well, isn’t that odd? Pec major does three different things! But let’s go on! If I start with my shoulder hyperextended (that is behind anatomical neutral), then pec major sternal (remember, the shoulder extensor?) now becomes a shoulder flexor. We could claim that the muscle only has one role, and that anyone who believes otherwise is wrong wrong wrong. Or we could accept that the initial conditions of the system determine how the muscle will function.
In simple point of fact, the only thing a muscle really does is get shorter. That is how it works. If a muscle has one goal in life, it is to bring its two insertion points closer together. With muscles that cross only one joint, this is pretty simple. Brachiallis at the elbow gets shorter and the elbow flexes. Biceps brachii, on the other hand, crosses the elbow and the shoulder, as well as spanning the space between radius and ulna. So when the two ends try to get closer together, there are many possibilities: The shoulder could flex; the elbow could flex; the radio-ulnar joint could supinate. To accentuate the desired effects, the brain fires synergists. So if elbow flexion is desired, brachiallis also fires. However, if shoulder flexion and supination are NOT desired, the muscular synergy must include a shoulder extensor to neutralize the shoulder flexion, and a pronator to neutralize the supination. The biceps is relatively simple compared to the psoas, but you can already see the layers of complexity in this simple muscle. To understand the many functions of the psoas, imagine 5 thread spools, connected with a string end to end through the middle, and these are attached with a hinge to the side of a sink, tilted on its side. Attached with a hinge to the bottom side of the sink is a stick. Now imagine you have a string connecting the top spool to the stick. If you tightened the string, would the stick move forward? would the spools? would the sink? Maybe the spools would collapse on one side while the sink moves backward and the stick stays still? Of course, the correct answer is ‘It depends’. If you could stabilize the spools in some way, or neutralize the movement of the stick, or if the spools started in front of the sink… all of these things and more contribute to the many potential functions of the psoas. The ‘furthermore’ in all of this is that people who produce more force in hip flexion have greater cross-sectional area of the psoas (Kubo et al, 2010, Copaver et al. 2012), hip flexion has been shown to be weakened after LIF surgery, which interferes with psoas but not iliacus function (Lee et al. 2013) and the psoas has been shown to contribute to hip flexion (Lewis et al. 2009) and both hip and lumbar flexion, as well as some other complex movements (Andersson et al. 2007). By contrast, Hu and others (2013) found that when people performed straight leg raises while standing on one leg, the psoas was active on both standing and gesture sides. They interpreted this to mean that maybe the psoas was only acting to stabilize the spine. Of course, maybe the muscle was stabilizing the spine on the standing side, but lifting the leg on the gesture side. On the balance of the evidence, I maintain that the psoas has many and varied roles, but to claim that it is not an ‘actual’ hip flexor is a rejection of the ‘actual’ scientific evidence. And thus endeth the rant. As well as my pedantry.
I don’t dislike her article, and I love Stu McGill, whom she quotes, but there are babies and there is bath water, and we have to make some distinctions…
Don’t get me wrong. Having looked over the website and Katy Bowman’s excellent credentials and thinking, I hope Katy has the opportunity to influence lots of thinking around anatomy and biomechanics, because she is clear, down to earth, and cares about stuff I also think is important. I just have a few quibbles with some things written in her post. I hope to add a couple of video examples to make an illustration, but that will have to wait.
Feel free to comment, if I can figure out how to turn that on…