The Jaw Joint

Put a fingertip just in front of your ear and open your mouth as wide as it goes. Partway through, something under your finger moves forward and drops away. That's the top of your lower jaw leaving its socket. The joint it moves in sits between the jaw and the skull, so it's called the temporomandibular joint, or TMJ. There's one in front of each ear, and the two always work together.

Here is the lower jaw, the mandible, with the upper teeth and the two see-through sockets it hangs from. Drag to turn it. The mouth is open ; drag that number, or the slider, to open it wider:

For the first part of the opening the jaw swings like a door on a hinge. Then the whole jaw starts moving forward as well: the rounded top of each side of the jaw, the condyle, leaves its socket and slides down a bump of bone in front of it. That forward slide is what you felt under your finger.

A hinge and a slide

To see what happens inside, slice through one joint from back to front and look at it from the side, with the face to the right. Bone the slice cuts through is hatched. Open the mouth and watch the condyle and the bone above it, the socket and the eminence in front:

For roughly the first 20 mm of opening, measured between the front teeth, the condyle stays in its socket, the fossa, and only turns (orange). Past that it starts to slide forward and down the articular eminence (green), and it keeps turning as it goes. Stop at 22 mm to catch the moment the slide begins. At full opening, usually 40 to 55 mm in adults, the condyle sits close to the top of the eminence.

Turning alone can't open the mouth very far. The front teeth are about 10 cm from the joint, so the 13° the jaw turns in the hinge phase moves them only about 22 mm apart. The slide adds the rest, and because the eminence slopes downward, it also lowers the whole jaw. The plot shows the two phases: flat while the jaw only turns, then climbing.

A hinge could open the mouth all the way on its own: turning the jaw about 30° instead of 13° moves the front teeth 50 mm apart. So why slide at all? Compare the two ways of opening and watch the back corner of the jaw as well as the front teeth:

Turning alone swings the back corner of the jaw, its angle, about 28 mm backward at full opening, into the soft tissue under the ear. With the slide added it moves about 18 mm for the same opening, because the whole jaw shifts forward as it turns. The joint wouldn't allow the pure hinge anyway. A band on the outside of the joint, the temporomandibular ligament, goes taut near the end of the hinge phase; to open further, the condyle has to move forward .Ligaments only limit motion; they don't drive it. The muscles do the moving, and the ligaments and the shape of the bone decide where the moving has to stop or change course.

The disc in between

The condyle doesn't press bone against bone. Between it and the socket sits a pad of dense, fibrous tissue called the articular disc. Behind the disc is soft retrodiscal tissue, and in front of it the lateral pterygoid muscle. Drag to turn the joint, use the slider to cut it open, and point at any part to read what it is:

The disc is thickest at the back, about 3 mm in the posterior band, and thinnest in the middle, about 1 mm in the intermediate zone. It splits the joint into two compartments, each sealed and filled with lubricating fluid. The lower one, between the condyle and the disc, handles the turning. The upper one, between the disc and the bone, handles the sliding.

The retrodiscal tissue behind the disc is the opposite of the disc's middle: soft, full of blood vessels and nerves, and painful when squeezed. The lateral pterygoid has two heads. The lower one pulls the condyle forward when you open; the upper one attaches to the front of the disc and the condyle and steadies them when you clench. The ear canal sits just behind the joint, which is one reason jaw pain is often felt as ear pain.

Its shape is easier to read with the disc lifted out and laid flat. Drag the probe along it, from the back on the left to the front on the right, and read the thickness:

The posterior band is about 3 mm thick, the intermediate zone about 1 mm, and the anterior band about 2 mm. That shape is most of what keeps the disc in place. The condyle sits in the thin middle with a thick rim in front and behind, the way a marble sits in a dent, and the pressure of the closed joint holds it there.

The attachments do the rest. Behind, the retrodiscal tissue has two layers: an upper layer with elastic fibers that runs to the skull, and an lower layer that runs to the back of the condyle. In front, the disc joins the joint capsule and the upper head of the lateral pterygoid. At its inner and outer edges, short ligaments fasten it to the two ends of the condyle, so it can tip forward and back on the condyle but not slide off sideways .

A disc that rides along

The joint only works if the thin middle of the disc stays between the two curved bone surfaces, where the gap is narrowest. Here is the same side cut as before, now with the disc and the soft tissue around it. Open the mouth again and watch the disc:

As the condyle slides forward, the disc slides with it. It also turns backward on top of the condyle, so that by full opening the thick back band sits behind the condyle instead of on top of it, and the thin middle is still in the narrowest part of the gap. The orange line marks the surface where the turning happens and the green line the surface where the sliding happens.

Behind the disc, the retrodiscal tissue stretches and fills the space the condyle leaves behind, its blood vessels swelling to take up the room. Its upper layer is elastic and helps pull the disc back as you close. In front, the lower head of the lateral pterygoid shortens as it pulls the condyle forward.

The muscles that move it

The joint decides where the jaw can go. The muscles decide where it goes. Three pairs close the jaw: the temporalis, the masseter and the medial pterygoid. The lateral pterygoid pulls the condyle forward, and a group under the chin led by the digastric pulls the jaw open. Turn the model and pick a movement to see which ones do the work:

The temporalis fans out over the side of the skull, above and in front of the ear, and narrows to a tendon that grips the coronoid process, the thin blade of bone at the front of the jaw's upright part. The masseter is the thick muscle of the cheek; put your fingers on your cheek and clench, and you'll feel it harden. The medial pterygoid mirrors it on the inside of the jaw, so the two hold the jaw's angle in a sling of muscle .

Each muscle pulls along a line, and the direction of that line is what it does. Here is the jaw from the side with every muscle drawn as its line of pull. The arrows show how hard each pulls on the jaw for the movement you pick:

The masseter and the medial pterygoid both pull up and a little forward, which is why they're so good at closing: their lines pass well in front of the joint. The front fibers of the temporalis pull nearly straight up, and its back fibers pull backward, which is how the jaw is drawn back after you push it forward. The lateral pterygoid is the odd one out. It runs almost horizontally from the condyle toward the middle of the head, so it slides the condyle forward down the eminence, both when you open and when you jut your chin out.

Opening takes much less force than closing. Gravity helps, the digastric and its neighbors under the chin pull the chin down and back, and muscles below the hyoid bone hold the hyoid still so they have something to pull against. Closing is where the strength is, and it's what the next section is about.

A lever with the joint at one end

Treat the jaw as a rigid bar. The closing muscles pull up on it somewhere in the middle, the food pushes back down on the teeth, and the joint holds the back end. Drag the bite point along the teeth, and watch how the muscles' pull is shared between the teeth and the joint:

Nothing moves, so the turning effects about the joint have to cancel. The muscles' pull \red{F_M} acts at a distance \red{d_M} from the joint, measured square to its line of pull; the bite force \green{F_B} acts at \green{d_B}:

In this model \red{d_M} is about 35 mm. A first molar is about 48 mm from the joint, so it gets about seven tenths of the pull: 577 N of an 800 N pull. The front teeth, 84 mm away, get about four tenths, 330 N. Turn on the lever arms to see both distances.

The rest of the pull doesn't disappear. Whatever the teeth don't take, the joint does: with a molar bite the condyle presses on the skull with about 310 N, and with a bite on the front teeth with about 500 N. Here are both forces for every bite point at once:

That's why you crack a nut with your back teeth without thinking about it, and why biting hard on something with the front teeth feels like a strain on the joint. Measured maximum bite forces vary a lot between people and with how they're measured, but they run to several hundred newtons at the molars and much less at the front .

Chewing

Chewing isn't a straight up and down. Watched from the front, the lower front teeth trace a loop on every stroke, and the loop leans toward the side the food is on. Pick a side, and use the speed button to slow it down:

Each stroke has the same four parts. The jaw drops, straight down or a little toward the other side (blue). It swings out toward the side with the food and starts to close (orange). It closes through the food on that side, crushing it between the back teeth (red), the only part of the stroke where much force is needed. Then the teeth glide together back to the middle . A stroke takes a little under a second. The loop drawn here, about 17 mm deep and 5 mm wide, is a common size; tough food makes it deeper and wider, soft food shallower.

Swinging the chin to one side uses the two joints differently. Look down at the jaw from above and move the chin:

On the side the chin moves toward, the working side, the condyle mostly turns in its socket. On the other side the condyle slides forward and down the eminence, the same move it makes when you open, and the jaw pivots around the working condyle. Push the chin straight forward and both condyles slide together. In a real jaw the working condyle also shifts a millimeter or so sideways, a movement dentists name after Norman Bennett, who described it; the model leaves that out.

Every place the jaw can go

Put the pieces together and ask a plain question: where can the front teeth go? Move the jaw with the pad. Sideways slides the condyles forward down the eminence; up and down opens and closes the mouth. The graph marks the point the lower front teeth reach, as seen from the side:

The reachable positions fill a tall, narrow shield. Its back edge is the hinge, then the hinge and slide together, all the way to the widest opening. Its front edge runs from the chin pushed fully forward down to the same widest point. Its top edge is set by the teeth: from teeth together, the lower front teeth have to slide down the back of the upper ones before they can move forward. Everything you do with your jaw, from talking to yawning, happens inside this outline, and most of it in a small region near the top.

Why jaws click

In a lot of people the disc doesn't sit where it should. It slips forward, so that with the mouth closed it lies in front of the condyle instead of on top of it. MRI studies of people with no jaw complaints at all find a displaced disc in roughly a third of them . Switch the disc between its normal and its displaced position, then open the mouth slowly and close it again:

With the mouth closed, the condyle now presses on the retrodiscal tissue, the part with the nerves, instead of on the disc. As it starts to slide, it runs into the thick back band of the disc and pushes it ahead, until it pops over the band onto the thin middle. That jump is the click. From then on the joint works normally, until near the end of closing, when the disc slips forward off the condyle again with a second, usually softer click.

The two clicks happen at different openings: the opening click comes later than the closing one. Clinicians call this disc displacement with reduction, because the disc is out of place but returns, or reduces, every time the mouth opens . On its own and without pain, it's common and usually left alone.

Locked shut

Sometimes the clicking stops, and not because the joint got better. If the disc stays in front of the condyle all the time, it becomes a doorstop. Try to open this jaw as far as the other one:

The condyle can still turn, so the mouth opens a little, but it can't slide past the disc, which bunches up against the eminence. Opening stops at around 25 to 30 mm, roughly the end of the hinge phase. Clinicians call this a closed lock, or disc displacement without reduction.

Because the jaw has a joint on each side, a lock on one side also bends the opening. Here are both joints from the front with the left one locked. The midline of the upper teeth is marked with a dashed line:

The right condyle slides forward normally while the left one stays in its socket, so the jaw swings around the stuck side and the chin moves toward it, here by about 4 mm to the left. A jaw that opens to one side, together with a limited opening, is one of the things clinicians look for when they suspect a closed lock. (The figure is seen from the front, so the patient's left is on your right.)

Locked open

The opposite problem happens at the far end of the range. With a very wide yawn or a long dental appointment, the condyle can travel all the way over the top of the eminence. Open past the normal range, then try to close:

Once the condyle is in front of the eminence, the muscles that close the mouth pull it up against the front slope of the bump (red arrow), and that holds it there. The harder you try to close, the more firmly it's held. The mouth stays stuck open. This is a dislocation of the jaw, sometimes called an open lock.

To put it back, a clinician presses down on the back teeth, which moves the condyle down below the top of the eminence, and then guides the jaw backward so the condyle slides into its socket. The button above does the same two moves.

Wear and clenching

Bone isn't fixed in shape. It remodels to suit the loads it carries, and the jaw joint carries a lot: every bite, and for many people long stretches of clenching or grinding, often during sleep, at forces like the ones in the lever above. Over years the condyle and the bone above it can change shape. Slide from a healthy condyle toward a heavily remodeled one; the drawing is an illustration, not a scan of any particular joint:

Some of this is adaptation: a joint that is loaded more grows a broader, flatter surface to spread the load, the same way other bones thicken where they're used. When the changes go further, with the cartilage-like surface breaking down, bone rubbing on bone and spurs of new bone forming at the edges, it's osteoarthritis, the same process that affects knees and hips. Long-standing disc displacement often comes with it, and the disc itself can thin or tear.

The link to symptoms is loose. Scans of people with no complaints often show remodeled condyles and displaced discs, and many people with jaw pain have joints that look normal; the pain then usually comes from the muscles, which can ache from clenching just as any muscle aches after holding a load for hours .

Pain that doesn't settle, a jaw that suddenly can't open more than about two finger widths, or one that's stuck open is a reason to see a dentist or doctor; the rest of this page describes how the joint works, not what to do about a particular one.

Two small joints

Every time you speak, chew, or yawn, two joints the size of a fingertip do something no other joint in the body does: they hinge and slide at once, on a pad that rides along between them, and the two of them, joined by one bone, have to agree. The hinge gets the mouth started, the slide finishes the job and keeps the jaw out of the throat, the disc keeps the bones from grinding, and the muscles, pulling along their lines, turn the jaw into a lever that cracks nuts at the back and takes care at the front. Most of what goes wrong, the clicks and the locks, is a small pad in the wrong place at the wrong time.

The figures are drawings built from one simplified model, not measurements of any one person. The joint is shown as a single side cut, the muscles as straight lines of pull, and the lever as a rigid bar in two dimensions. Sizes and angles are typical adult values, rounded.

References

1. Jeffrey P. Okeson, Management of Temporomandibular Disorders and Occlusion, 8th edition (Elsevier, 2019). The standard clinical textbook on how the joint is built, moves, and fails.
2. Susan Standring (editor), Gray's Anatomy: The Anatomical Basis of Clinical Practice, 42nd edition (Elsevier, 2020), the chapter on the infratemporal region and the temporomandibular joint.
3. Ulf Posselt, "Studies in the mobility of the human mandible", Acta Odontologica Scandinavica 10, supplement 10 (1952).
4. William L. Hylander, "The human mandible: lever or link?", American Journal of Physical Anthropology 43 (1975), 227–242.
5. William L. Hylander, "An experimental analysis of temporomandibular joint reaction force in macaques", American Journal of Physical Anthropology 51 (1979), 433–456.
6. Duygu Koc, Arife Dogan, and Bulent Bek, "Bite force and influential factors on bite force measurements: a literature review", European Journal of Dentistry 4 (2010), 223–232.
7. Richard W. Katzberg and colleagues, "Anatomic disorders of the temporomandibular joint disc in asymptomatic subjects", Journal of Oral and Maxillofacial Surgery 54 (1996), 147–153.
8. Eric Schiffman and colleagues, "Diagnostic Criteria for Temporomandibular Disorders (DC/TMD) for clinical and research applications", Journal of Oral & Facial Pain and Headache 28 (2014), 6–27. The definitions of disc displacement used on this page.