Dog breeds are grouped — often arbitrarily or erroneously — into from five to ten categories based on function, superficial appearance, or geographical origin, depending on the registry organization. Just because it may make more sense to assign them to groups based primarily on ancestry and then on historical function, does not mean that such will be the case. In most dog circles, the “working” breeds have always been considered as those that originally did such work as herding or guarding livestock, pulling loads, and protecting property. Even though other breeds had specific occupations in the service of man, they are not known as working breeds: sighthounds running down prey or predators, gundogs flushing food for the table, terriers and toys terrorizing vermin — these were more or less doing what they would do without human ownership, anyway, so their jobs were considered less like “work”.

Many dog organizations split the huge Working Group into two, with the ones that had historical development for tending, driving, or bunching flocks and herds being called “Herding breeds”.  Never mind the confusion about whether the reindeer-herding Samoyed is hardly much different from the sled-pulling Husky — that’s a puzzle for another time. Most of the Group that did not resemble the mastino-type wagon puller or the bear-fighting wooly flock guardian type were once employed to trot around the animals raised by man for his food, and assigned to the herding subcategory. These latter were specialists in trotting, in covering much ground with the most efficiency (least effort). This meant that success favored those with the most shoulder angulation over those with the stiffer, more vertical front ends.

 When we speak of a shoulder in a dog, we usually include a lot more than just the scapula (shoulder blade) — although the flat, broad bone is often the center of attention. No part of a dog exists alone, not even those “floating” bones such as the hyoid, sesamoid, clavicle, patella and penile bones, all of which are connected to muscles and other bones by ligaments and tendons. The shoulder is intimately related to most other portions of the foreassembly or “forequarters”, from the skull to the ribs, from vertebrae to arm and breastbone.

At the most forward and lowest portion of the scapula is a shallow socket in which articulates the head of the upper arm (humerus). This area, especially the humeral greater tubercle that protrudes in front of the articulation, is called the “point of the shoulder”. Running roughly up the center of the blade from that point nearly to the top is a ridge of bone known as the spine of the scapula. The lowest and thickest section of that spine close to the socket is the acromion. See Figure 2. There are several muscle groups attached to the scapula. Don’t be frightened by their names; use abbreviated nicknames if you stumble over pronunciation. What matters is that you know what action each gives to the blade and the entire limb, and where the muscles are attached.

The first of these is the triangular trapezius muscle originating on the bones and ligaments of the vertebral column from the third cervical (neck) vertebra to the ninth thoracic vertebrae. Its insertion is on the spine of the scapula. Since part of this thin broad muscle lies forward of the ridge it is attached to, and part extends to the rear, it can easily be seen that its function is primarily to elevate the limb. It also brings the arm forward and helps in changing the angle of layback during movement. The omotransversarius starts from the first cervical vertebra (the atlas) next to the back of the skull, dips beneath the other muscles of the neck, which extend to the sternum and arm, then attaches to the scapular spine near the acromion. Its obvious action is to draw the limb forward and rotate the bottom of the scapula forward while other muscles are trying to hold the rest of it in place. Knowing that, you can easily understand why a dog trotting in the show ring isn’t going to extend its forelimbs in ideal or equal reach if it has its head turned toward its handler. Nor will it cover as much ground if it trots with its head held high instead of forward and slightly above back level. Yet terrier and cocker dog-show handlers are notorious for the silly spectacle in which their dogs often barely touch the floor with their front paws!

Beneath the trapezius lies the rhomboideus muscle, which originates on the vertebrae from near the head to about the sixth or seventh thoracic vertebra. Its insertion is along the edge of the scapula, farthest from the acromion. Because of its wide origin, it can lift the limb upward, pull the limb and shoulder forward or backward and draw the scapula against the rib cage, depending on which portions are ennervated (stimulated by nerve impulses.)

The muscle filling the space in front of the scapular spine is called the supraspinatus, and it is attached to the top of the humerus. Thus, you can envision it straightening out the shoulder-arm angle and bringing the limb forward. The (behind the spine) will either flex or extend the shoulder joint, depending on the position when the muscle contracts. It also is inserted on the humerus. Other muscles include the infraspinatus, serratus, teres, deltoideus, and sub-scapularis; all play some parts in moving the scapula in relation to the ribs, vertebrae or upper arm.

What is “layback?” See Figure 1. It is the front-to-back inclination of the shoulder blade, seen and felt when one touches both the point of the shoulder and the top of the scapula or the withers at the same time. The withers is the area atop the shoulder from where the neck ends to where the “true back” begins. In most dogs, the last cervical vertebrae and the first thoracic vertebrae are down between the shoulder blades, so you might not be able to feel them. The withers is thus a transition stage between the neck’s relatively upright carriage and the nearly level back called for in most working breeds. (I use the term “working” in the utilitarian sense, and especially the herding dogs.) The beginning judge (or the one evaluating heavily-coated breeds) often checks and compares layback by running his thumbs down the spine of the scapula.  That line is almost parallel to, and only an inch or so behind, the envisioned line from the highest point of scapula to point of shoulder.

A very unfortunate situation has arisen out of ignorance and laziness: many AKC breed standards were written by people who copied the wording from other erroneous standards without checking accuracy first. The same problem is seen in the multitude of breed books in which artists’ drawings of canine skeletons could almost have been traced from other books, judging from the mistakes they have repeated.

I had been preaching scientific honesty and artistic accuracy for years, thinking I was like Elijah: “the only prophet in the land of Baal”, until Rachel Page Elliot’s book Dogsteps came out. As I had done on a smaller scale, she x-rayed many standing and running dogs to prove the nonsense about 90 degree angles between humerus (upper arm) and scapula, and the impossibility of a 45-degree layback of the shoulder.

The Bouvier’s AKC standard has quite good wording, as does the Collie’s. The AKC German Shepherd’s “about a right angle” is misleading. Is slightly less as good as slightly more?  Is it even possible? The Dobe standard is a travesty, what with the 45-degree layback and 90-degree shoulder/arm angles being specified. Might as well specify cubical tires for cars! That so-called right angle cannot even be approached if you draw your lines down the middle of the upper arm (if you have x-ray vision, maybe you can tell where the center line of this slightly curved, well-padded bone is!).

If you draw your line from point of shoulder to the highest part of the ulna that we call the point of the elbow (leaving the humerus to do so), you get points of reference you can see and feel. Now draw your second imaginary line from point of shoulder to top of shoulder blade. The angle between is closer to 90 degrees than if you tried to use the line going through the shaft of the humerus, but you still don’t get a right angle, even with the best laid-back shoulder blades. In my live-dog illustrated lecture, “Analytical Approach to Evaluating Dogs”, I show where these lines are by drawing chalk lines on dark, short-coated canine volunteers. Even when I don’t use a protractor, my audiences can see the fallacies of those printed standard specifications. See Figures 1 and 4.

Why are some shoulder angles better than others? In a few cases this is the same as asking why some scapulas are laid back at a greater angle than others, though most deficiencies lie in the upper arm rather than the scapula layback. If the angle of the spine of the scapula does indeed differ between dogs, it is possibly because some dogs have proportionately shorter vertebrae in the neck or sacrum, and some may have longer bones in the true back and loin (the true back being between the scapula and the croup). If the dog has shorter vertebrae and disks, the shoulder may more upright.

Remember that those trapezius and rhomboideus muscles extend to the ninth and sixth thoracic vertebrae with only a small portion of their fibers, and if the vertebrae are relatively short or the scapula is slightly more upright in that dog for another reason, the muscle attachment will be relatively forward and less broad. The same may be true to a lesser degree with the serratus ventralis, which runs from the scapula to the last five cervical vertebrae and the first seven ribs. The more forward all these attachments are, the less the blade will be inclined when a pup begins to move and muscle forces help shape its semi-cartilaginous bones and joints. In such a dog, there would be less muscle mass present to rotate the top of the scapula back and forth, thus a restriction of motion here contributes to a lack of reach in front and even follow-through beneath.

There is a far greater cause for poor reach, or less smoothness of front action. Some dogs have an upper arm (humerus) that is not laid back at a good angle from point of shoulder to elbow joint. This is sometimes accompanied by a proportionately shorter arm compared to the scapular length. Such dogs are in the minority, but it’s wise to keep an eye on the problem. Because of cyclic neglect, German Shepherd Dogs, have periodically become alternately better and worse in this respect. At the time of this revision, the AKC-GSDCA type has lacked good upper arm layback for many years while the international type has improved since the 1970s.  See Figure 3 for one artist’s conception of ideal structure.

Whether herding livestock, doing police work, performing obedience exercises, or pulling loads, the working dog needs a well-angled shoulder/upper arm assembly. Let’s consider this synonymous with good layback of both bones, for convenience’s sake. A “straight” (more vertical) foreassembly is somewhat like a car without springs. Imagine a dog with poor front angulation hitting the ground with its forelimbs after climbing over a wall in pursuit of an errant lamb or thief. The hard shock will have a detrimental effect before long.

A dog with better angles (yet strong ligaments in pasterns, elbows, and shoulders) can spread that shock over an imperceptibly longer period of time, during which the muscles slow the impact while the bones go through their “folding up” action relative to each other, then release that stored energy by straightening out again (bouncing back). Trotting creates very nearly the same sort of shock that jumping does, only far less violent.

A successful parachutist survives because he takes only a tiny fraction of a second longer to hit the ground than someone whose chute didn’t open. A good boxer “rolls with the punches”, while the guy who holds his head still when the other guy’s fist approaches finds himself waking up some time later. The baseball player relies on padding and moving his hand back to slow the speed of the ball as it makes contact with his glove. The differences in time intervals in each illustration are truly minute, but they can mean the difference between ease and pain, or life and death. Likewise the differences in layback from dog to dog may be small, but a tiny difference can mean smoother action, greater ability to hit the ground effortlessly whether jumping or trotting, and a longer useful working life.

Not many dogs are used for pulling loads anymore, but the dog with a smaller angle between scapula and upper arm is better suited for this type of work, too. Even if only for historical interest, the ability to pull carts or sleds should be preserved in those breeds that are developed for such purposes, for breed type is inextricably bound to that utility. Form follows function, and if we get too far in the evolution of breeds from their original purposes, we will have created (though gradually) a distinctly different breed. What time traveler from centuries past would recognize today’s utterly non-functional English Bulldog from those he had seen chase and tame wild bulls in the days when the breed had a useful purpose? Do not let our working, utilitarian breeds slip away into uselessness as some other breeds have. There is a good reason why I put so much emphasis on shoulder and upper arm angles, and reward good examples in the show ring. It’s the same reason why it is so difficult to improve in breeding.

In the case of angulation at the knee (“bend of stifle”), ignorance and fad-following have resulted in GSDs with lower-thighs that are too long, with hocks too far behind the torso to be adequately controlled by ligaments and muscles. This rear angulation at the stifle can go either direction from the middle (moderate) ideal for most breeds, with the American-fashion GSD at one extreme and Chows at the other. However, in the case of the shoulder, the ideal is not in the middle of a normal range, but on one end; namely, closer to the fictional 45-degree layback and 90-degree shoulder/arm angles. Actually, depending upon breed and whether you draw the bottom line to the point of the elbow, a layback of 35 degrees in either scapula or upper arm is very good, and a withers - point of shoulder - point of elbow angle of 95 degrees is excellent for a herding breed. Drawing that line through the upper arm, you would get about 115 degrees in most of the efficient trotters. If it were possible to create a 45-degree layback in both bones, some say that such a dog might fall on its face.

But back to my point. And let’s consider the palpable points of reference to be the indentation at the point of shoulder, the highest point of the scapula, and the top of the ulna. If a number of genes affect the angle between these bones, some would be “ideal genes” (let’s say they’d call for a 35-degree angle from vertical, for each bone, for purpose of illustration), and others would be “less than ideal genes” calling for some lesser angle. Of all the possible genes that could be transmitted, the vast majority would be calling for an angle of somewhere in the 20 degree to 30 degree range, with a miniscule number calling for the coveted approximately 120 degrees remaining between the scapula and humerus. Genes are inherited randomly, and statistically would show a bell curve with the smallest amplitude in the two extremes (say, 15 degrees and 35 degrees, for example) and the greatest in the middle of the curve. Despite all the talk about angles, it boils down to this advice: forget the numbers, examine as many dogs as you can get your hands and eyes on, compare one dog to the next, and reward or admire those with the smallest apparent angle between shoulder and upper arm, while proving what appears in stance by watching the dog perform in the trot. See Figures 4 and 5.

Since he cannot do “better” than the ideal shoulder angle, which is at one end of a range of possibilities, the breeder must be more diligent in such an instance to cull from breeding programs all dogs which drift an undesirable distance from that good end of that spectrum. More so than is necessary in any trait in which the ideal is at some intermediate point between the worst on one end and the worst in the opposite direction. In the case of good forequarter function in a herding breed, and in most other working breeds, there is only one direction from the ideal, when we speak of breeds developed for trotting.

To paraphrase Sir Edmund Burke, eternal diligence is the price of freedom from poor forequarters.

Fig.1  Shoulder Angles – It is almost impossible to duplicate, by eye or hand, the typical illustration in most books that shows a 90-degree angle between limbs, with lines going through the middle of the humerus and from either the most-forward point of the shoulder or the imagined location of the center of articulation to the highest point of the scapula or along the scapular spine. Only in the “best” fore-assemblies will an angle of 90 degrees even be approached, and then only if lines are drawn on radiographs from top of ulna to front of upper arm to a point behind the highest point of the scapula.

Figure 3  Various Bone-Joint Angles in a Well-built Herding/Working Breed

(Actually, few GSDs have this good a shoulder, and very few from American lines since the 1970s)