Training From Scratch (Special) – What muscle does that work?

This post applies the Analytic Fitness™ methodology to one of the most common and yet often useless questions in strength training in order to determine when it’s worth bothering.

“What muscle does that work?” is an interesting fitness question asked daily by countless casual gym goers who have no business considering it in the first place and that countless trainers keep answering to rather than explaining to their clients why they have no business considering it.

Reasons abound for this situation, both sad and bad. For the sad ones, most trainers actually don’t know they have no business considering the question in the first place and therefore don’t know their clients have no business asking it either. That’s probably because basic physiology is usually the only science personal trainers can impress their clients with have a minimal education in.

Unfortunately, it takes little energy to drop a few Latin words in the conversation. Much less than it takes to attempt to educate someone about motor patterns and neural pathways, or the neural component of strength. Especially if that someone is going to ask you the question again after the lecture anyway. So why bother with the lecture in the first place?

And that’s how trainers actually know enough biomechanics and neurology to understand that the question is largely irrelevant for most people end up answering it anyway. And yet, only a modicum of biomechanics and neurology suffices to:

  • dispel the need for asking this question save for a few cases where special circumstances arise that would make it relevant, and:
  • understand when those special circumstances arise and make the question relevant.

This post is an exercise in Analytic Fitness™ that illustrates the above, and then some. It proceeds as follows: first, I discuss why almost everybody who is making a fuss about it shouldn’t even bother with the question; and then why it’s worth raising an eyebrow when actually nobody does.

Why is everybody making a fuss?


To put it simply, the question “What muscle does that work?” is virtually irrelevant for strength sports or any serious physical activity that involves resistance training.

Whether resistance is provided by gravity (gymnastics), the environment (swimming, rowing), a combination of both (rucking, mountaineering) or equipment (cycling, any strength sport), counteracting a resistance is a whole-body thing. Well, maybe not with cycling, where it’s just a whole-limb thing. But that’s it.

(To anybody is tempted to wisecrack something like: “Wait, ain’t rucking just walking with a backpack? That’s a leg thing too, innit?” I’ll just say one thing: wait until the next “Old School Strength” post.)

Even if one considers only strength sports broadly construed (Olympic weightlifting, powerlifting, strong(wo)man, and kettlebell lifting) individual muscles or well-identified muscle groups do not matter as much as motor patterns. On occasion, try to ask the question to any serious strength athlete in one of those sports about their competition movement, and they will most likely answer ‘all of them‘. If they answer at all.

Not from sports

If the question does not come from sports, where does it come from then? The short answer is bodybuilding.

What muscle does that work?” makes some sense in the context of bodybuilding ‘body part splits’ with boatloads of muscle-specific isolation exercises. Unless a sport is seriously imbalanced there is little point to consider training body parts. Again, bicycling would be an example, but that’s it.

The thing is, isolation exercises are mostly useless for muscle-building in the absence of performance-enhancing drugs. So the long answer is about twice as long as the short answer: enhanced bodybuilding.

The most popular performance-enhancing drugs among bodybuilders and other enhanced athletes are androgenic anabolic steroids (AAS). They contribute to elevating protein biosynthesis, a pedantic way to refer to the ability of the body to use amino acids broken down from food to building muscle tissue (and the linings of blood vessels) all which was already covered in this post. The effect of AAS is systemic, a mildly pedantic way to say that it occurs all across the body and is largely independent of the type of exercise.

In fact, study published in 1996 found that supraphysiologic injections of testosterone without training stimulus were more efficient than a well-designed exercise program without injections for building muscle and nearly as effective for building strength. Greg Nuckols published an excellent discussion of that study on Strengtheory Stronger by Science a while back so I’ll just refer you to it if you want details. Otherwise, the plot below is self-explanatory for the purposes of this discussion.

Steroids with and without exercise

The plot illustrates is that enhanced trainees can rely on ASS for muscle growth and use isolation exercises as a resource-management tool helping their body prioritize specific areas. In fact, enhanced trainees probably could as well do isometric holds to a similar effect (incidentally, that why isometrics have unjustly fallen in disrepute).

Strength training without drugs

By contrast, non-enhanced trainees can only stimulate short-term local protein synthesis in response to exercise. Furthermore, muscle mass is metabolically costly to maintain. Without anabolic steroids to reduce maintenance costs, non-enhanced trainees quickly lose their understimulated muscle tissue. The loss can be rather fast if the body needs the resources for something else (like building the cardiovascular system).

Stimulation through exercise lasts between 24 and 48 hrs, non-enhanced trainees cannot make long-term progress with body part splits and often regress when switching to them from whole-body routines. (Christian Thibaudeau has recently addressed this topic in T-Nation a lot beginning with this post.) The body-part-split-plus-cardio regimen that most non-enhanced trainees stick too in the hope of looking like bodybuilders who train like that is just what prevent them to look like they lift. But their VO2max is probably higher than average, which is not that bad.

In most circumstances, the best exercises for non-enhanced trainees are those for which the answer is as close as possible to ‘all of them. That’s clearly the case when training for strength, and especially for strength sports, but it also holds true for hypertrophy. As for strength, given the principle of specificity, training sports movements or close variations of those movements is the best way to train sport-specific strength. And since sports movements involve “all the muscles” the close variations are going to involve them as well.

That being said, training for a strength sport is not the best way to build an aesthetic physique, irrespective of enhancement. Protocols that maximize performance in sports do not typically promote muscle growth as well as other types of mechanical work. Isolation exercises have their place in non-enhanced bodybuilding provided that frequency is adequate. But the frequency has to be ramped up, both for the training sessions and for the number of times one trains each body part.

This routine from Christian Thibaudeau illustrates this perfectly. Based on a summary of the clarifications offered in the post I mentioned already, Thibaudeau proposes a plan based on 6 training sessions/week where body parts are trained 3 times/week. He also suggests how to adjust it for 5 days/week or even 4 days/ week, but in the latter case his go-to program for size includes only two exercises, and they both work “all the muscles”.

Christian Thibaudeau training all the muscles

What everybody should really be making a fuss about

The question”what muscle does that work?” has some relevance outside of the narrow context of drug-enhanced cosmetic muscle-building. But in those other contexts it has almost nothing to do with muscle-building.

Let’s backtrack a bit and re-consider the question itself, as it is in fact not entirely unambiguous. Applying a common distinction from analytic philosophy, we can interpret it as either a normative or a descriptive question. Hand-waving the details of the normative vs. descriptive distinction, it boils down to this in our present case:

  • Normative/General: What muscles should contribute to that movement under normal circumstances, when a fully-functional human being performs it?
  • Descriptive/Case-based: What muscles do in fact contribute to that movement under the specific circumstances of me or you performing that movement?

We need to attend to some details about the normative vs. descriptive distinction for the special case under consideration, so let’s get to it.

Normative vs. Descriptive 101

In the mind of most philosophers, the normative vs. descriptive distinction is associated with the name of David Hume. Interestingly, in the context of muscle activation, another distinction of Hume’s applies, namely the distinction between ‘relations of ideas’ and ‘matters of fact’. But this being an Analytic Fitness™ post, and not an analytic philosophy post, I’ll leave the theory for another day, and just illustrate it. Simply put, the first questions is about relations of ideas and the second about matters of facts. Here’s why.

The first question is about an abstraction (the ‘normal’ human). In order to answer it, we try and draw logical consequences from our knowledge of physiology and biomechanics as they apply to that abstraction. Furthermore, both physiology and biomechanics are reasoning from abstractions:

  • Physiology looks at the origins and insertions of muscles in our skeleton and hypothesizes what movement should result from the contraction of said muscles.
  • biomechanics looks at the mechanical forces that our structure as a whole must counteract in order to maintain its integrity while performing a given movement, then hypothesizes which muscle should contribute to maintaining said integrity.

It’s not pure guesswork because it’s based on data, but since it generalizes from the data, it’s still educated guesswork.

The second question is about matters of fact (you and me, under particular circumstances). In order to answer it we would need empirical data putting into practice our knowledge of physiology and biomechanics. Now, there are no ‘naked’ facts, so this always involves a bit of theorizing and hypothesizing. But it also requires the observation of actual subjects (you and me) who may have developed abnormal movement habits.

Before moving forward, let’s note that physiology and biomechanics agree most of the time but sometimes they don’t. To put it simply, that’s because biomechanics is more ‘holistic’ than physiology. If you are curious about cases of and causes for disagreement, you can check Stuart McGill’s Low Back Disorders: Evidence-Based Prevention and Rehabilitation 3E(Chap. 3: “Functional Anatomy of the Lumbar Spine”). I’ll come back to the topic in the future. For now, let’s move on to an example.


A common pathology is gluteal amnesia which happens when the gluteal muscles — gluteus maximus, medius, and minimus (cf. picture) — fail to activate normally during leg and hip extension. This pathology was investigated by Stuart McGill (again) who credits Dr. Vladimir Janda for the suggestion that it would be widespread (cf. Low Back Disorders, p. 150). The inhibition of gluteal muscles is most likely a consequence of prolonged sitting as it entails both a lack of use of those muscles and cumulative trauma due to their compression.

Failure to activate the gluteal muscles can happen during both normal locomotion or exercises that target the glutes. Other muscles that are not inhibited by sitting will then take over, like the quadratus lumborum (QL). The QL also compensates for other muscles inhibited by sitting (like the erector spinae) and tends to be overworked as a result. In combination with gluteal amnesia, hyperactive QLs are a common cause of lower back pain (cf. McGill’s Low Back Disorders 3E [pp. 85-86]; if you don’t own a copy, you can check the Wikipedia entry).

In a study published in 2009, Stuart McGill (damn, him again!) and his collaborators found the demand imposed by a heavy yoke walk on the gluteal muscles could exceed their maximal power output. How could strong(wo)men managed to complete successfully an ‘impossible’ task? Well, the co-contraction of the trunk musculature required to maintain spinal structural integrity permits to recruit the QL to develop the additional power needed to move the load.

If you are not familiar with the yoke walk, here’s a demonstration below by Kristyn Whisman, multiple time winner of America’s Strongest Woman, carrying 290 kg (636 lbs) for just under 10 meters (30 ft).

So, interestingly, the QL can become involved both when the gluteal muscles do and don’t work the way they should. work the way they should. Hence, the same motor pattern can result from pathological or positive adaptations: here, the recruitment of the QL to assist with gait either inhibited gluteal muscles in daily life, or fully functional gluteal muscle in need of assistance during athletic exercise.

So, what muscle does walking work? Well, it depends. If you’re a fully-functional human, the glutes and the quads. If you’re a chronic sitter, the quads and the QLs. If you’re a fully functional strong(wo)man, the glutes, the quads, and the QLs (possibly even when you’re not under a yoke, due to neural adaptations, but that’s another topic).

And in case you wonder about stron(wo)men with sleepy glutes, well, they probably won’t walk very long (a topic for another day).

A glimpse at the longer story

Trying to answer to the question”what muscle does that work?” is, after all, an interesting exercise, but it’s more complicated than looking at an exercise description in some exercise library.

Without getting into too many details, the main issue is with how the central nervous system (CNS) manages muscle activity. The CNS keeps some muscles under tension even at rest, while letting others alternate between phases of rhythmic contraction and phases of relaxation. Based on this difference, Vladimir Janda (mentioned earlier) suggested differentiating muscles as tonic (tension at rest is called muscle tone) and phasic muscles.

The differences between phasic and tonic muscles go a long way to account for the difference between normative and descriptive answers to out titular question. Phasic muscles that relax too often lose the ability to properly contract when they should: they tend to become inhibited. Tonic muscles, who never really relax, and are then left to (literally) pick up the slack, becoming hypertonic. The conditions by which a hypertonic muscle takes over the function of an inhibited muscle is called compensation.

Let’s just take the special case of pathological adaptations to sitting. The muscle map below shows the tonic muscles and the phasic muscles that are respectively shortened (contracted) and lengthened (relaxed) when we sit. Eventually, imbalances and compensations in the neck, lower back, shoulders and around elbows and knees will happen, turning into chronic discomfort or pain.

The relation between muscle imbalances and chronic pain is the basis of the Janda approach. And as McGill later showed, chronic pain results in non-normal muscle activation during motion. This turns the question “what muscle does that work?” into a tricky one as soon as some of the compensation has to get factored in.

Of course, this only applies if to people who sit a lot. But if that’s you (chances are that it is) most of the standard description of exercise libraries are inaccurate as far as you are concerned. They take physiology for granted and assume a ‘normal’ case while neglection neglect:

  • biomechanical constraints that may impose that some muscle groups ‘off-load’ the work on others even under normal circumstances; and:
  • pathological adaptations to injuries or cumulative trauma leading hypertonic muscles to take over inhibited ones.


Strength athletes are usually more aware of biomechanical constraints than casual gym-goers and that’s probably why they turned the mind-muscle connection biceps-curl skinny guy Internet meme.

Unfortunately, strength athletes also sit a lot and are particularly exposed to injury. So, they probably should also pay more attention to constraints that result from pathological adaptations to sitting (which are not always weeded out by rehabilitation, a topic in future posts).

Now, to make of all this?

Well, it depends on whether you are considering the whole training-from-scratch-thing hypothetically or not.

In the first case, you have hopefully emptied one more cup of tea (assuming if it needed emptying in the first place), corrected some misconceptions, and perhaps become aware of a possible source of performance issues (repeated sitting) that was not high on your priority list. Congratulations.

In the second case, it gives you a basis to set a priority list: forget about muscles, think about movements, and remember that you suffer from pathological adaptations to sitting that you must correct before you address anything else.

All of which will come in handy to design a training-from-scratch program. If your interest was hypothetical, you will nonetheless be able to use the program as a remedial one. And if it was not, you will understand its design.

But that would be a topic for another day.


This post is a Patreon-exclusive version of a very polite, very civilized post that I wrote some time ago for but I have since pulled out. The differences are: (1) some style editing; (2) additional information that did not make it in the original post for length reasons; and: (3) tie-ins to other posts on

Leave a Reply