Since there are cave passages (and cavers) of all shapes and sizes, there are a host of different ways of moving underground, some of which are essentially unique to the cave environment. This entry is an attempt to give an overview of the different methods, excluding technical and vertical methods (rock climbing, ladders and rope work). Some odd situations require peculiar ways of moving, as the following story illustrates... This researcher recently had a couple of trips to the end of an old mine level a mile long, quite narrow, between 5 and 6 feet high, flooded to about a foot deep, and with an uneven rubble-strewn floor and submerged railway tracks. By experimenting and comparing speed with cavers ahead who were walking relatively normally, albeit stooped, it became clear that it was actually faster to walk in a stooped semi-goose-step, bending and lifting the trailing leg up clear of the water, straightening it out whilst swinging it as far forward as possible, then overbalancing onto it. This helped to avoid many of the brick-sized rocks and railway sleepers on the floor, and involved less dragging of the feet through the water than is the case for normal walking. Eventually, however, the unnaturalness of this gait required a return to more standard walking; but, for speed and variety, it was definitely useful.
Depending on the presence and consistency of mud, the angle and smoothness of the rock and the nature of footwear, grip underground on sloping surfaces can vary from nonexistent to awesome. Unlike damp surface rocks, there is no slippery layer of algae, and scrambling on steep, clean-washed rock underground can be a blissful experience. In the UK, Wellington boots are the standard caving footwear, though people who have not tried climbing in wellies often don't appreciate the quality of grip that a good pair can provide.
Each kind of movement has its own requirements in terms of strength and energy, which vary with the shape, strength and fitness of the caver and the speed of movement. Generally, the easiest method and speed for a given passage will be chosen, but circumstances such as the depth of water in damp passages or a desire to progress rapidly1 may require the use of a less comfortable or more tiring motion.
Even where the ideal method varies constantly due to fluctuating passage dimensions, it is sometimes easier to maintain one gait than be constantly changing. For example, there is little point standing up from a crawl to walk a few paces in a higher section before dropping back into a crawl, or temporarily climbing down from a traverse to take a short walk along the passage floor before climbing back up again. Conversely, in a long passage of consistent shape, it may sometimes be desirable to switch from the method of first resort to give tired muscle groups a temporary rest, or simply to relieve the monotony.
Hazards also have to be taken into account - loose rock, height, water and excessive constriction should provide pause for thought. In some caves and particularly in old mines, there can be places where you really don't want to risk the most gentle of touches on the walls or roof. Even for horizontal movement, if there is significant vertical exposure and a risk of falling then traverse lines2 should be used for protection.
Bipedal Floor-level Movement
Given that cave passages all have walls and that much of the time one or both of them are within reach, it is quite common for the arms to be used to aid balance and assist in changes of direction and forward propulsion even while walking or running.
Where the passage is wide enough, and the floor is not too slippery, this is the fastest method of progression. In the case where the passage floor is uneven due to the pattern of erosion or the presence of fallen boulders, but there is not sufficient vertical variation to be dangerous, dynamic movement by rock-hopping can be safe and efficient. Skipping along from one high point to the next can avoid repeated tiring cycles of stepping or climbing down-and-up, and the increased stride length possible with dynamic movement can increase the choice of footholds. Avoiding down/up climbing this way also reduces the effort required from the arms, which has a definite benefit in that for most cavers, the arms fatigue more quickly than the legs.
Where possible, a caver will walk, since it is both a comfortable way of moving and requires little effort, especially when compared to the various forms of crawling.
When a passage is sufficiently tall, but is too narrow for regular walking - which can be a common situation in sinuous Yorkshire canyon passages - it may be possible to keep walking in a roughly normal fashion, but with the upper body rotated to face across the passage, using the arms for balance, or if the walls are sufficiently smooth, giving arms a rest by sliding one's back along a wall. If the lower passage width does not permit the legs to move past each other normally, an alternating legs apart/legs together movement may still be a possibility. In the case of extreme narrowness at floor level, the feet may have to point along the passage while stepping sideways. Alternatively, where a reasonably wide passage tapers down at floor level to a foot-trapping width or less, it may be necessary to traverse along with feet pointing across the passage supported by heels and toes on opposing walls some way above floor level.
If the usable height of the passageway is less than that of the caver, walking can drift into intermittent or continuous stooping. In narrow passages, the walls can be used for support, whereas in passages of sufficient width, the extreme lowest-height form of stooping involves partially bent legs, a roughly horizontal torso pointing across the passageway, and broadly sideways or diagonal movement.
Such sideways progress enables the head to be turned to the side, allowing both the roof and floor ahead to be seen, whereas a head-first posture would require the head to be bent right back for forward vision, which is less comfortable on the neck and which would require the upper body to be bent over slightly more. Moving sideways, or semi-sideways, can also provide more choice of foot placement on uneven floors, and improved clearance over floor obstacles. Hooking the trailing arm under the trailing leg can help to keep the body bent without straining back muscles, and aids lifting the leg those crucial extra inches.
Quadrupedal Floor-level Movement
When a passage is too low even for stooping to be feasible, it is necessary to use a four-limbed motion to progress. The tallest quadrupedal stance uses hands and feet on the floor to carry the caver's weight. Humans are not natural quadrupeds, and our arms are not sufficiently long to maintain this posture for long with any degree of comfort, so it is generally restricted to short dips between higher sections of passage. The main exception to this rule is the case where a passage too low for stooping has a sloping floor in cross-section, usually where there is a bank of sediment sloping up from the floor to one or both walls. In such a passage, a sideways movement with feet on the floor and hands on the bank can be comfortable and fast. Given enough room, arms and legs may scissor past each other in motion, and even a bounding sideways movement is an option. Additionally, if the caver is moving through a suitably low passage while dragging a heavy tackle bag, attached by cord to the caver's waist, the rearward balance provided by the bag can allow hand-and-foot progress with less strain on the arms than usual.
The next-tallest method is the classic hands-and-knees (or forearms-and-knees) crawl, which with the use of kneepads3 and a somewhat smooth floor can be quite comfortable even when carried out over long distances.
One crucial factor in determining possible positions is the length of the upper leg, since if this is greater than the passage height then the knees cannot be brought forwards under the hips, and so the legs must therefore be splayed out, and swung forwards to the side of the body if the passage width is sufficient. This reptilian gait can be more tiring than a straight-ahead movement, but has the advantage that it is possible to elevate the entire lower leg above the floor and place it down carefully at the end of the forward motion, which can be much more comfortable on a sharp or uneven floor than a standard crawl, where the knee is at a greater risk of hitting bumps on the floor when being advanced.
As passage height continues to decrease, the position shifts to the use of forearms and lower legs as the weight supports, commonly with the forearms lying sideways or diagonally across the passage. The lower (and narrower) the passage, the straighter-back the legs become, and the more difficult it becomes to use them for propulsion.
In low passages, it may be possible to progress while keeping the torso off the floor by supporting the weight on toe-tips and forearms, and holding the body and legs rigid in roughly that same way one would for press-ups. Slow propulsion can come from a combination of moving the arms and 'walking' on the toes; or faster movement can be gathered from the arms alone, dragging the rest of the body along on the toes. This requires a reasonably smooth floor and appropriate footwear.
Flat-out crawling is the next stage, and whether on chest or back (or on the side in more vertical slot-like passages), in the absence of suitable purchase for hands or feet, the muscles of the torso can be used in wriggling motions to assist progression. If lying on the back, usually where the arms are being used to manoeuvre a tackle bag or where the floor is very smooth yet there are some handholds in the ceiling, the shoulders can be surprisingly effective for movement over short distances.
There are many circumstances when the floor of a passage is of little use for progression. Possibly the bottom of a rift is too wet or narrow to be comfortably passable, yet the passage is wider at a higher level. There may be a climb up or down in a passage, yet the climb may be difficult to make at the obvious point, yet easier somewhere further along the passage. On occasion, the passage floor drops away but it is desirable to move forwards while maintaining height either to reach a continuation of the current floor level, or to reach the best point to descend a pitch4.
In these circumstances, it is usual to traverse - to move more or less horizontally above floor level. In contrast to many surface canyons, which have a tendency to widen consistently from bottom to top due to surface weathering processes, in underground canyon passages, the width at any given height depends essentially on the precise composition of the rock and the historical water flow at the time when the rock at that particular height was being eroded. It is commonly found that subtle variations in the composition of the limestone beds and the varying developmental history of the cave results in variations in the vertical cross-section of the passage that tend to be maintained to some degree along the length of the passage. Put simply, it is common for there to be sets of ledges throughout the height of a rift passage (or at least of narrower and wider areas) that persist horizontally (or along the dip of the limestone beds) for some distance.
Where to Traverse
Choosing the level (and hence width of passageway) at which to traverse is most important - too narrow a section may be strenuous, and especially awkward when carrying a tackle bag, whereas somewhere too wide, or high, may be excessively dangerous as far as falls are concerned.
The width of the passage has a huge influence on the ease and possibility of traversing, and on the particular methods that are practical, as do the nature of any available footholds/handholds, the smoothness of the rock, and the presence of lubricating mud. Grip to resist the pull of gravity is obtained by a combination of the use of protruding holds and friction, between the caver and rock surface. Increased sideways force can increase friction, and this can either be actively provided by muscular force - generally where holds are limited and the rock is mud-lubricated or very smooth - or by the caver's own weight, acting at an appropriate angle. Even a sloping foothold can provide good grip if the caver is leaning at an angle across the passageway, or has splayed legs, provided that the force transmitted down the leg is at such an angle that it tends to keep the foot on the hold.
Even quite smooth vertical walls of a suitable spacing can provide support, given strenuous effort, possibly with feet on one wall, and hands and backside or shoulders on the other. However, the inability to rest in such a situation and the high risk of slipping limits such moves to areas where the risk of injury is low, due either to limited risk of falling or the use of rope protection.
At the widest extreme, it may be necessary to progress with feet on one wall and hands on the other, with the body leaning over face-down at quite a shallow angle. This kind of move should only be attempted when quite confident, since a slip could easily result in a head-first pitching down the passage if traverse lines are not being used for safety.
In a similar way to floor-level running, allowing a different choice of moves compared to walking, when traversing on good ledges with limited exposure the dynamic balance available from rapid movement, can sometimes allow movement using much less upper-body effort than would be the case for slower moves. For example, if a passage roughly 1 metre wide has ledges 10cm or more, deep on either side, even if those ledges slope downwards to some extent, it may be possible to bounce along the passage alternating from one ledge to the other with little or no use of the arms, whereas a slower movement would require serious lateral support from the arms for balance. Effectively, the sideways force needed to prevent the body keeling over into the rift, and possibly to improve grip on a sloping ledge, comes from the sideways momentum imparted to the body by the previous jump-off from the opposite ledge.
Similarly, in a smooth-walled passage, or one with hand and footholds too small to use individually, it may be possible to maintain horizontal momentum while switching support between hands and feet alternately braced across the passage. That is, one move involves pushing out sideways with the feet on opposite walls, while rotating the body forwards over the pivot of the feet, with the other move being support of the body by hands on opposite walls while the body swings forwards under this new pivot. Rapid alternation between hands and feet can allow a larger force to be used to improve friction against the rock without tiring as rapidly as would be the case where the same force were applied for longer in a static situation.
Dynamic vs Static Movement
Some moves are only possible and others are made easier, by moving at speed. This is relatively obvious for the case of running, boulder-hopping or jumping across gaps, but is also true in other cases. If forced to stoop for a short section of lowered roof, moving at speed can help improve the balance compared to moving more slowly. Whether or not it is actually the case, the feeling from some dynamic moves is that tendons are absorbing and re-releasing energy and the sensation of distance-eating progress can help boost flagging morale. Additionally, if a gait such as a stooping walk is uncomfortable, getting it over with as soon as possible can be a good thing. In particularly tiring situations, like traversing between smooth walls, it can be as energetic to stay still as to move, so the natural temptation to stop and 'rest' needs to be moderated by thought.
The problem with some dynamic moves is the requirement for confidence - changing one's mind about when to jump to another rock or bounce to the opposite ledge while traversing can make an easy move into a painful failure. It is always worth assessing alternative possibilities, such as which 'backup' holds or places to aim for in the case of a slip, yet it is best to keep the good moves one intends to do at the forefront of the mind.
It is rather like walking on a narrow path round a steep mountain slope - someone who keeps thinking 'If I slip, I'll fall down there, or even 'I must avoid falling down there' has the thought '...fall down there' rolling around in their mind, and runs a small risk that their body will do just that if pushed into quick action by a minor crisis such as a stumble. Someone who keeps thinking 'If anything happens, I'll dive into the uphill slope and grab on to something' has a reasonable chance that their reflex action in case of anything happening will be to scramble to safety.
With experience, and given the immense amount and variety of visual and tactile feedback available while caving with full awareness, a caver can develop a very accurate 'body image', knowing precisely how large they and their equipment are in all directions and various kinds of attitudes. Indeed, it is possible for the caver's 'body image' to become so finely tuned and automatic that simply switching to a slightly taller helmet can result in days of glancing blows from low ceilings before corrections are made. Once aware of the options, a caver then needs to figure out which is the best of the available methods of movement.
Combining that strong sense of self with the skill of looking at a passageway of constantly fluctuating dimensions, to see exactly where and how they could move, can become second nature. However, in awkward or unusual situations there may still be the need for much thought and experimentation.