Animal Combat Loops

After years of research, I can say with confidence that object-based combat is the basic criteria for hominization, which then unfolds into grammatical language, and use of language causes self-domestication.

Object-based combat is exclusive to humans. We see chimps use objects but only for foraging, as in nut-cracking, and predation, as in using sticks to fish for ants and termites. We also see crows, otters, and other animals using objects in non-combat scenarios.

Chimps are rarely seen wielding objects during combat, but it’s so infrequent that, in all my reading on the similarities between chimps and humans, I can think of no times when paleontologists have mentioned them. But when chimps do wield objects in combat, such object-usage is not part of the combat loop.

Animal Combat Loops Described

I’m borrowing from video game development terminology when I say that chimps never use objects in their combat loop, nor do any other animals. The combat loop is defined as the mode of attacking, defending, posturing, and retreating within any intra-specific combat (chimp vs. chimp, buck vs. buck, etc.). Combat loops with animals are therefore highly restricted, being limited to their natural weapons, natural defenses, communication signals, and means of fleeing.

It’s worth watching some animal combat loops for context, if only to marvel at the techniques. Here are some bucks fighting. Once they lock up, the footwork is super intricate, and there’s clearly a lot of touch sensation when determining when and how to make the next move:

Here’s a fight between hippos where teeth are the only useful weapon at play. It looks like the one on the right has an old puncture wound on its hide. The technique appears to be to scoop under the jaw:

Here’s a bear fight that goes on for over 9 minutes. It’s interesting to watch how wrestling plays out when biting is both permitted and useful. There’s a lot going on technique-wise, but it’s most surprising that the main weapon in play is the teeth. I don’t imagine the claws do much to their skin:

And here are lions fighting. This combat loop resembles bear combat, only since the paws are less articulated there’s increased emphasis on teeth and agility:

Animal combat loops are lengthy but rarely end in death. It’s hard to know whether the animals want to kill one another, but if it were true, we’d expect them to kill opponents that submit. This doesn’t appear to happen within the same species. The “submit” pose appears to terminate the combat loop, almost without fail.

Animal Scapegoating

However, animal combat can and does end in death, as in the Gombe “Civil War” among chimps. I wrote about this phenomena here. We see plenty of examples of animals ruthlessly scapegoating their own kind, which is fuel for the sentiment that “humans are just animals” while ignoring the obvious fact that humans use objects for scapegoating and will inadvertently draw others into the scapegoating act.

Rene Girard’s adherents would have us believe that scapegoating itself is at the foundation of human society, but no attempt is made to differentiate human scapegoating from animal scapegoating. Or worse, Girardians might not even acknowledge animal scapegoating exists, or might call it something other than “scapegoating.” The fact that animal scapegoating does occur and yet does not produce complex cultural forms within the species demonstrates that scapegoating itself is not the minimum requirement for hominization.

I’ve also made the note many times within the Generative Anthropology community that the threat of apocalyptic violence at the Originary Scene which creates the first linguistic sign needs to meet a certain threshold of violence by some means. This was never outlined in GA literature (which is otherwise very good, and I encourage you to read more here). I propose that object-based combat provides the minimal necessary conditions which demand a sign be produced.

Because without object-based combat, without the threat of objects as weapons, apocalyptic violence is impossible, otherwise we would see it in animals, which we don’t. The animal combat loop is perfectly compatible with scapegoating, which should not be treated as an anomaly.

Anomalies in Animal Combat Loops

Other anomalies can arise in animal combat loops. Bears might fight until they fall off cliffs. It might be an honest mistake, neither of them accounting for the environmental conditions:

We also see chimps fighting on cliffs and not giving it much attention. Note that the attackers never attempt to push the one chimp off the cliff, nor does the latter try and get away from the cliff, or hop down to safety (the full video shows the drop to be ~20 feet, hardly a lethal drop for a chimp):

As I’ve shown before, a chimp uses a stick as a weapon, but when it’s lost he doesn’t give it an additional thought, and we might assume his technique wasn’t emulated.

Aside from these, anomalies are extremely rare and non-repeatable within animal combat loops. Whatever combat loop we see between members of a single species is most likely the same combat loop that existed at the inception of that species.

Animal Combat Loops are Optimized

As I’ve mentioned elsewhere, animal combat is optimized because the combat loop is closed. Anomalies are non-repeatable. Humans could introduce weapons into animal combat and it wouldn’t make a difference to the overall shape of combat.

All combat loops generate a field between the antagonists, a shared moment of anticipation, which I call Merge. In Merge, signals are verified, rejected, ignored, etc., producing a vocabulary. When Merge is shallow, as in optimized combat loops, there’s little time for signals to develop, and little need: the optimized combat loop sets the hierarchy efficiently and without threatening the species. Animals will willingly enter combat loops because risk of death is relatively low. Animals also back down more willingly because the cower signal is practically guaranteed to end the fight.

Signals like ears pulled back, tail down, snarling, and growling have shared meaning between the antagonists, verified by Merge. Animal antagonists can be certain that, so long as the combat loop never changes, then the Merge field will not change, and the same signals will be accepted by Merge. And that’s exactly what we see: animal signals are simple, quickly shared, and understood by both parties. Animals tend not to randomly grow new weapons. If a cat suddenly grew a horn, such a development would change the Merge state, which would demand new signals be Parsed to figure out what new signals can be shared thanks to the addition of this horn. After a few combat loops the lexicon would be stable, at least until a new weapon is added.

Animals can be certain (if they are certain of anything) that their combat loops will not change any time soon. If they do, the changes will be very gradual, like an extra toe appearing on a cat, or a different looking eye. Since combat loops remain the same, Merge remains the same, and the lexicon is locked in place. We can expect that the neurological wiring is efficient enough so that the animal can learn this lexicon easily, and there won’t be any wiring necessary to learn beyond this.

Unoptimized Combat Loops in Humans

The human combat loop is infinitely more complex than the animal combat loop. Because both human antagonists have to 1) estimate what the opponent’s weapon is going to be while also 2) anticipating that the opponent is estimating what Ego’s weapon is going to be. This creates an infinite feedback loop, which is an infinitely deep (or recursive) Merge field, which never really closes, even if/when the antagonists part ways for fear of mutual annihilation. This always-open 24-7 Merge field makes human combat unoptimized because the hierarchy can’t be set so easily as in animal combat. Both antagonists realize that, if they submit, the opponent might suspect they are just going to get a new weapon later, or bring some armed kinsmen, so submission loses its lexical value, unless there’s some guarantee that the submitted foe will remain submitted.

The recursive human Merge field therefore demands that, for submission to be accepted as a legal signal, this term must be Parsed in a new way to guarantee that no retribution will come. All of this demands innovative lexical items (words, gestures, etc.) which compound the vocabulary. Not only are there new terms, but the terms themselves are burdened with the infinitely deep, recursive properties of the human Merge field, and they themselves can modify other lexical items. So human communication is suddenly infinitely deeper than animal communication, with each signal standing for the recursive Merge state. In other words, each human word can infinitely modify other words. The signal for “I give up” now must be paired with other terms like “and here’s proof” or “or God strike me dead.” Or the signal for “I have a rock” must be paired with “but I don’t want to use it.” These words all have their own in-built hierarchies. “Rock” can be modified with “my” or “your” or “grey,” but not with “happy.” “Food” can be modified with “fragrant” or “share” but not with “thunder.”

These simple compound terms seem obvious now but were hardly obvious at the beginning of human language. Such combinations would have been tried in the Merge field which would have rejected most of them, only Parsing those that confirmed the structure of the Merge state. Those terms and their relationships to one another can be tested away from the field of battle since human, recursive Merge never closes. The vocabulary is allowed to grow to whatever size is necessary, and the grammar can be as complex as possible.

Due to the wildcard nature of human violence, the combat loop is totally different not only for each individual conflict, but even within individual conflicts themselves. If the fight suddenly moves into a river, the water is now a weapon. If it moves to a plateau, now the cliff is a weapon. Merge is therefore constantly changing, which demands reevaluation of the language during every conflict and within the conflicts themselves.

Human grammar is thus a product of the unoptimized human combat loop, which creates a recursive Merge state, and every lexical item/word inherits this depth. It’s why every language’s grammar is relatively the same in its depth. Every one of them has verbs, nouns, direct objects, pronouns, and a host of other grammatical categories. Some have articles, like the English “a, an, the” but this is hardly evidence of advancement over a language like Chinese which has no articles, since it simply solves this specificity problem with more context. Cantonese has a host of final markers to denote a sentence’s tone, which are absent from English, but perhaps we employ more body language and accents to convey the same tones.

An interesting study would be to examine the grammars of stone age cultures such as the Ganowanian family of Native American languages or the Australian aborigines and compare them to Iron Age languages like Latin, or further looking at the development of High German grammar after the usage of gunpowder in warfare. Did these grammars shift and become more complex due to the sudden change in the major mode of aggression? Maybe someone can chime in. Whatever this development looked like, it’s possible that this is what put pressure on the brain, fueling its growth as the lexicon grew, not in number of words, but in the hierarchical depth that each word brought.

In conclusion, animal combat loops create stable Merge states that basically never change, and animals therefore don’t have grammar. Human combat loops are unoptimized and always changing, creating infinitely deep Recursive Merge states that are different in every interaction, which demand complex Parsing schemes that constantly introduce new terms while verifying old terms and throwing out outdated ones, and this is why we have grammar.