Dealing with knife attacks

[Warning: this article contains a graphic image which might distress some readers.]


In modern times it has become fashionable to regard the traditional martial arts as ineffective against knife attacks. And yet if you look at the techniques in traditional forms, I believe you will find plenty of realistic knife defences.

This should not be surprising: traditional fighting arts are, after all, principally civilian defence systems.

And throughout history, in every culture, knife and dagger attacks have remained relatively common in civilian society - precisely because knives (or similar bladed tools) are so ubiquitous: We have used them in our daily lives for thousands of years, be it in hunting, skinning/scaling and gutting, cooking as well as general utility.  (There is a reason that the Swiss Army Knife is first and foremost a knife - not a screwdriver, saw or file etc.)

At the same time, the knife has enormous advantages as an instrument of civilian (not military) attack: it is small and portable, can be easily concealed and it is the next most deadly civilian weapon after a firearm (as I discuss later).

So I believe the traditional martial arts have evolved very much with knife attacks in mind.

The problems we see with traditional knife defences don't lie with the traditional technology - but rather with the way this technology has been interpreted in a modern age when it is relatively rarely needed and applied.

Before we look at specific, traditional knife defences, it is appropriate that we look briefly at what knife attacks really look like.

What you can realistically expect in a knife attack

Many years ago I was involved in at least 3 knife attack prosecutions which included surveillance footage evidence (I was involved in many more that did not have such footage).  I can confirm that what most people still practise as "knife attack scenarios" in the traditional arts are woefully inaccurate and that knife attacks typically take a form much more like the one below:

Let's look at the features of the above attack scenario:

The full step zombie attack vs reality

Your attacker won't take a full step and pause like some sort of zombie, giving you time to control the arm holding the knife, disarm your opponent and inflict a counter.

Your attacker is very unlikely to allow you to lead his/her momentum as you often see in aikido demonstrations.

Rather, I think it is most likely that the attacker will, after closing the gap, lunge at you with repeated short stabs (and maybe hacks and slashes - see below).

These movements will be fast and continuous.  They might be straight, but given most people's instinct to cover their centre line, they will probably move around your guard (eg. roundhouse stabs to the kidneys).

The "Psycho" hack and the "slash"

I think the "Psycho" attack using the downward hack, while certainly possible, isn't nearly as prevalent as people seem to think.

The "Psycho" attack - it's not very realistic!

While the Jim Carrey video above is just a parody, it is self-evident that many schools of traditional martial arts teach the downward "hack" (in zombie mode!) as the principle attack one can expect to face.  Yes, you could well be attacked with a downward hack.  But in my view this possibility comes second to the more common straight or round stab - executed with a fast retraction.  

You would not think this if you went by what many people in traditional schools (even highly reputable and esteemed ones) demonstrate as their first "defence against a knife attack".

From what I've seen and prosecuted, most attackers (certainly in Western countries) don't bother much with slashes either (unless the attack is specifically intended to scar victims - eg. what is known as the "Glasgow smile" - a very different sort of attack from the type I'm considering here).

Certainly I cannot remember a single knife prosecution which involved slash wounds to any appreciable extent.

Yes, experienced Filipino martial arts fighters use slashes and hacks extensively (see Guro Dan Inosanto below).  But this does not appear to be the case in most "non blade" cultures.

It is important to note that even when a hack or slash is employed instead of a stab, the same principles apply - to both the attack and defence.  You can expect a frenzied rush - and a very mobile, fast, knife-wielding hand.  You certainly won't face a "Jim Carrey" attack.  

This greatly influences what you should and shouldn't do.

What you shouldn't do

Given the above, whatever you do, you shouldn't aim to grab your opponent's knife hand - particularly as it is heading out to you.  This is just plain daft.

First, the hand will be travelling too fast.  Grabbing a moving hand at speed is a big ask - even with a punch, as Bas Rutten points out below:

Consider for a moment that if a punch is hard to "catch" in its outward phase, a knife stab is going to be even harder.  Why?

Punches are really far more likely to be launched with committed momentum than knife attacks.  A knife doesn't require anywhere near the commitment and momentum to do damage.  Moreover, the attacker will be very wary of having his main weapon caught and immobilised.

In other words, there is a greater premium than ever on keeping the attacking hand mobile when it is holding a knife.

Accordingly, if you try to grab a stabbing hand straight off the bat, you'll almost certainly miss.  What else will happen?  You'll probably get cut - severely.  Putting your hand near a fast moving blade is a very bad idea, as Guy Mezger found out when he tried to "block" a knife attack by raising his hand.  All the usual "blocking" principles that might work for boxing, MMA etc., in particular using the palm to deflect attacks, go out the window.

You just can't afford to use your hand against your opponent's fastest moving part - his/her hand - particularly when there is a very sharp blade in it!  You'll fail to get any sort of control and you will almost certainly get cut.

(Despite getting your hand in the way of the blade, you might, like Guy Mezger, still succeed in taking out your opponent.  But whichever way it goes, trying to grab the knife hand from the outset scarcely makes for an intelligent tactic - not that this was what Guy was trying to do).

While "grabbing the knife hand" isn't a great idea, you would likely get the opposite impression if you looked at most martial arts demonstrations of knife defences.

The traditional arts I've already shown above (karate and aikido) aren't the only ones who do this: you'll see it as far afield as Russian Systema.  Take a look below at one of Systema's teachers instructing a class on knife defences/disarms.

Sure, it  looks good.  But note the speed at which they are moving - a speed that disguises the difficulty of avoiding contact with the blade.  Look also at the nature of the attacks.  Compare this with the more realistic (in my view) attack scenario shown in the first video (stabbing, retracting the hand, etc.).

Please note: I am not saying that the Systema approach shown above wouldn't work well with "unbladed" fighting.  My own experience of Systema suggests that it is highly functional.  Such slow practise is also highly valuable in learning how to move smoothly and efficiently in relation to an attack.

However I hold it to be manifest that what is shown in the above video simply isn't appropriate for knife fighting.  As with many other schools/systems, the instructor seems to have fallen into the same trap: an unquestioned adoption of "unbladed" tactics in a bladed environment.  You can't do a simple swap: it just doesn't work.

What you should do

Okay, so what should you do?  You should do what I believe most traditional systems actually teach in their forms (as opposed to how they are interpreted).

The very first of these is quite simple:
Stop the slow moving parts first!
This is a basic principle of knife fighting.  Don't go for the fast moving hands: go for the shoulders and the forearm - preferably both at the same time.  Why?  Because even if you do nothing else, when you control these parts (ie. "cut the supply lines" as Marc MacYoung calls it), the best the knife can do is make a relatively shallow graze or cut.  It can't stab deeply.  You don't need tremendous force to stab, but you do need enough.  Having your shoulder and forearm jammed and controlled makes it very difficult to generate enough force.

When I was prosecuting, I noted that people who were stabbed deeply had the potential of dying quite quickly.  Major organs could be badly damaged, breathing could be interrupted and there was a whole lot more chance of bleeding from major arteries and veins.  Superficial slashes and cuts, while painful and sometimes nasty, just didn't compare in terms of mortality.

The vital issue of range

In order to stop the slow moving parts, you need to move in towards your opponent.  You can't do anything from a distance.  You have to be well inside the melee range.  Either that, or you should stay well outside the melee range.  Anything in between is "no man's land" and it is functionally useless, as well as being highly dangerous.

In other words, you must be inside the fast hand speed range or outside it - not in the middle of it!

Some traditional techniques: taijiquan

So let's consider some techniques from traditional arts.  Somewhat counter intuitively, I'm going to focus on the Chinese internal art of taijiquan - an art not normally thought of as containing defences against knife attacks:

First, let us examine the technique known as  "fair lady works at shuttles".  It is often interpreted as an interception and strike to the face.  But if your opponent is armed with a knife, it can be equally employed as an interception of your opponent's forearm and simultaneous jam of his/her shoulder.

Indeed, this is precisely the principle used in the Filipino martial arts of arnis/escrima/kali.  The adjacent picture is taken of me in taiji class, but it might just as well come from a Filipino martial arts class because it demonstrates the same basic principle:
Stop the slow moving parts first.

You'll note that the lead hand jams the shoulder, while rear hand jams the forearm in what I like to call the "Goldilocks zone".

Okay, I know this technique is against a "Psycho" type hack, and I know I said it wasn't the most common means of knife attack.  But such attacks are still quite possible.  Moreover, I've started with this example because it illustrates the essential principle very clearly: stop the slow moving parts first.  Then, and only then, do you attempt to counter and/or control your opponent's knife hand.

The need for a quick counter

Desmond Lawrence Sensei - one of my
esteemed seniors and teachers
In terms of "controlling the knife hand", my senior Sensei Des gave me the advice that rather than attempt to control the knife hand, it is often better simply to counter immediately.

This is what he did when attacked by 3 knife attackers on the docks in Cape Town.  He sustained 9 stab wounds (I've seen the scars).  But he knocked out 2 of his assailants while the other ran away.

Des then dragged himself to hospital.  He lost a lot of blood, but he survived.  The way he described it, the attack was simultaneous and frenzied.  He could do nothing but "wear" the stabs, punching each opponent as he came into range.

Indeed, Des' tactic wasn't dissimilar to that of Guy Mezger who took out his opponent by punching him: it's just that Des didn't even bother trying to stop the attacks.  He told me he would have tried to do so, but with 3 men attacking simultaneously this simply wasn't an option.

Controlling the knife hand after "cutting the supply lines"

However, it is certainly okay to attempt to control the knife hand if you are able to do so and if it is appropriate in the circumstances.  The important thing is to remember to stop the slow moving parts first!  Once you do so, two things happen:

1. You thwart the initial attack at its source by "cutting the supply lines", buying valuable time to respond.
2. You establish a kinaesthetic "link" - providing feedback as to your opponent's position in time and space.

Armed with the kinaesthetic data provided by the initial contact, and by the fact that you've interrupted your opponent's momentum momentarily, you now have a far better chance of grabbing his or her attacking arm safely - assuming it is necessary or prudent to do so.

Alternately you might throw a disabling counter - or just run away!

I discuss all of the above in my video below concerning two other taijiquan defences against knife attacks:

What taijiquan doesn't advise

If you search Google or Youtube with the terms "taijiquan" and "knife defense" you'll find many examples that contravene the basic principles to which I've referred above.  Some of these are from highly esteemed, greatly knowledgeable masters of the internal arts.

That you would get such results is hardly surprising: as esteemed as these masters might be, when it comes to knife fighting most of them are working largely, if not entirely, off theory alone.

This is not a criticism of these masters: thankfully, very few people per capita in the developed world have ever faced, or will ever face, a knife attack.  However it does mean that, in respect of knife fighting, the value of the information available is not particularly reliable.

This stands in contrast to the knowledge of our ancestors - who created the traditional forms during a time when knife fighting (indeed, any violence of an elemental, hand-to-hand kind) was far more likely to intrude into daily life.

As I've noted before, theory that is perfectly fine for unarmed fighting makes a very poor transition to bladed fighting.  Yet, time and time again, I see this manifesting in "knife defence" demonstrations from the traditional arts.  This is quite common and understandable.  I know that I've been guilty of it over the years.

Consider for a moment the adjacent series of pictures taken from the video below.  What do you notice?

First, there is a "zombie" attack, launched with a full step from out of range.  Even with the full step, the knife would barely reach the target: there would certainly be very little penetration.  To have any penetration, the attacker's arm would have to be fully extended.  This is simply unrealistic.  (Compare in my video my invitation to my student to come in closer when attacking).

Second, the defender's first response is to catch the defender's bladed hand.  As I note above, this is simply not feasible against any knife attack you're actually likely to face.  Even if you succeed in grabbing hold of the knife in such an application, very little awareness seems to exist in the traditional martial arts world of how easily the knife can be twisted on its axis to cut your grabbing hand.

Third, the advice offered in this video is to stay out of range.  As I've noted above, this is good advice.  But if you want to stay out of range, don't dive into "no man's land" to tackle the knife where it is travelling its fastest.  Just avoid it altogether!

Put another way: why intercept something that is going to miss you anyway - particularly if that thing is sharp and dangerous and you're intercepting the thing with your bare hand?  Look carefully at the above pictures and you'll notice that once the teacher steps back out of range, the knife is at least an arm's length away from him.  In that circumstance, why would you put your arm out into the path of the knife?

Even when the range is correct and the knife is controlled, very little attention is given to the fact that likely resistance and adjusted trajectory will have to be considered carefully.  Sure, you might have grabbed the knife hand - but this doesn't mean you're safe!

Consider the adjacent picture and note the paused position during the counter.  The attacker might not be able to reverse the knife trajectory or pull it back towards him.  But what stops him from changing the trajectory so that it continues downwards and sideways into the defender's groin or the inside of her thigh?  Not much, I think.  The defender is in a very weak position as regards that particular trajectory, yet she seems totally oblivious to this.


Knife fighting is very dangerous business.  While studying criminology we learned that knives were the next most deadly offensive weapon after the firearm.  Forget batons, nunchakus, knuckle dusters, baseball bats, broken bottles, half bricks, etc: after the firearm, the knife is the next most likely to result in death.

Because of the unique features of the knife, the tactics of knife fighting differ greatly from unarmed fighting or fighting with other "unbladed" weapons.

Most people reading this no doubt live in  modern world where, despite impressions they might get of rates of violent crime, their chances of being involved in an attack are statistically small - and their chances of facing a knife attack are even smaller.

This means that even well-intentioned, highly knowledgeable martial arts teachers are likely to "get it wrong" when it comes to knives.  This includes me. Even though I've seen and analysed knife fights in my professional career, I have no direct experience of being attacked with a knife (thankfully).  My seniors like Sensei Des (and even my brother) have, but I haven't.  So while I invite you to consider my opinion, I advise you not to take it uncritically as "truth".

Most of us are working from "theory".  I would like to think my theory is based on evidence, particularly evidence relevant to the society in which I live.  But I might be wrong.  This is why I haven't tried to deal with knife fighting as a subject before and why I approach it now with great caution, even after detailed study (including training in Filipino arnis/escrima) and careful consideration over many years.

In the end, the above internal arts teacher's advice is certainly spot-on in this respect:
Avoid knife fighting as much as possible!  
As to the rest, take most knife defence advice, including mine, with a very liberal pinch of salt...

Copyright © 2013 Dejan Djurdjevic