Many of you have probably heard of the hyperactive agency detection device, or ‘HADD’. It’s a great little concept that is incredibly useful to invoke when talking about numerous weird human beliefs. Why do we see faces in clouds? HADD. Why do some people believe in ghosts? HADD. Fairies? HADD. Conspiracy theories? HADD. Aliens? HADD. God? HADD. Got a question about some mysterious goings on that people seem desperate to attribute to the activity of some sentient being or other? HADD it mate. Throw HADD at it, case closed, problem solved! But what is the history of this ubiquitous HADD, and where is the evidence that HADD exists?
The hyperactive agency detection device has its roots in evolutionary psychology, quite a specific approach to evolutionary psychology in fact. Put very simply, the argument goes a little something like this. The human brain is almost like a Swiss army knife. At birth the brain is fully kitted out with pre-existing mechanisms or devices. Given the right environmental input, these mechanisms will spring into action at the appropriate time. These mechanisms have been passed through generation after generation, because individuals in our ancestral environments possessing these traits were more likely to successfully solve challenges of survival and reproduction, and therefore pass on these traits to offspring. Those offspring, endowed with these inherited mechanisms similarly did a cracking job at surviving long enough to create offspring. As long as the same challenge reoccurs, the same inherited solution is useful. Rinse, hump, repeat.
Hyperactive agency detection device is often argued to be one of those nifty little mechanisms that helped ancestral humans navigate a dangerous world. In a world full of lions and tigers and bears (oh my!), better to assume that the rustling sound in the bush is a hungry predator and run like stink, than to ignore it and become something’s lunch. So what if you occasionally leg it because the wind made a noise? Being a bit out of breath is far better than the awful, no good, terrible consequences you might face if you never run from the scary sound.
This logic of the hyperactive agency detection device sits comfortably within the larger framework of error management theory – the idea that some mistakes are far more costly than other mistakes. A classic example used to illustrate the importance of error management is that of the smoke detector. Would you prefer to have a smoke detector that is a touch too sensitive and causes a frantic outbreak of window opening, tea-towel flapping and exasperated swearing every time you dare to make toast, or would you prefer a smoke detector that sits placidly on the wall as your furniture and favourite trinkets get a lovely crispy crust. Clearly, you want the hysterical smoke detector that will take any excuse to sing its ear splitting siren song – better that than sofa flambé. The same is true for human cognition: if a false positive is annoying but a false negative is deadly, it’s best to calibrate your brain in favour of the false positive.
This idea, that humans are hyperactive agency detectors because assuming agents where there are none is less costly than not detecting agents where there are some, is beautiful. It’s elegant and charming in its simplicity and explanatory power. It makes intuitive sense, and it’s just so damn logical! There is just one tiny problem: there is no empirical evidence that the hyperactive agency detection device exists.
In Disbelief by Will Gervais, a few pages are dedicated to this specific problem. Although the concept of a hyperactive agency detection device has been around since the 1990’s, there is no evidence it exists. But this concept is so pervasive, especially among those who work in the cognitive science of religion, that it is almost treated as established fact. During my PhD I tried to investigate a link between heightened sensitivity to agents and religious belief and found… nothing. During their postgraduate days, Will Gervais and fellow researcher Aiyana Willard also tried to investigate the link between HADD and religion but found no evidence that HADD even existed.
After giving a talk at a conference detailing research that turned up no evidence that HADD was even a thing, Willard was approached by multiple other researchers who similarly spent time conducting research based on the idea that the HADD was the key to understanding religious cognition from an evolutionary perspective only to find the hyperactive agency detection devise was ironically undetectable. I have heard it said that researching a hypothesis generated on the assumption that HADD not only exists but is central to explaining religious cognition, then coming away with null results and the sudden need to change direction as quickly as possible, is tantamount to a right of passage for those of us who have spent time working in the Cognitive Science of Religion. The publication bias against null results is responsible for many doctoral tears.
But if HADD doesn’t exist, why is there so much anecdotal evidence that it does? Why do so many people have the experience of hearing a creak in the night and suddenly becoming filled with the fear that there is a burglar in the house? Why is HADD such an attractive, versatile and intuitively appealing idea?
In the paper agency detection is unnecessary in the explanation of religious belief Willard argues that our tendency to suspect a sentient being (or agent) is present is entirely dependent on context. We do not just assume any unexpected noise in any context or circumstance is an agent, but it depends entirely on previous experience, prior learning, expectations and likelihood. The paper opens by illustrating this beautifully:
A bump in the night can make us fear that a burglar is in the house […] If that bump is heard deep in the Canadian wilderness ,we are more likely to think – and are better served by thinking – that the sound is a bear. If we are skiing across snowy mountain peaks, an unexpected noise should make us fear an avalanche rather than any type of agent.
The likelihood that you will perceive a strange noise as a ghost is dependent on whether you live in a culture where ghost beliefs are common. If your culture is full of mischievous pixies, you are likely to attribute unexpected noises to mischievous pixies. The inferences you make will be dependent on the beliefs you already hold. So yes, many of us have had the experience of hearing an unsettling noise and thinking “burglar!” or being in a spooky place, hearing an unexpected whistle of wind and thinking “GHOST!” not because we have a hyperactive agency detection devise, but because of the expectations we have been endowed with due to cultural learning.
So, why is this story interesting? Two main reasons, I think. Firstly, that HADD is so often invoked, means it is important that we as skeptics are aware of the shaky foundations on which it sits. Secondly, and most importantly to me, it illustrates both the biggest problems and biggest potential in evolutionary psychology and adjacent disciplines. A common criticism of evolutionary psychology is that those who work in the area often tell very nice, plausible stories about the human mind with ideas that seem very logical but are ultimately unfalsifiable. The story of HADD and the research around it is showing something quite different. Yes, the idea has taken hold because it is a nice, pleasingly logical, easy to follow story, but it did allow researchers who are currently still working in the field of human evolutionary behavioural sciences (including evolutionary psychology) to generate falsifiable hypothesis and conduct research investigating this very idea. The idea is probably wrong, HADD probably doesn’t exist, and now researchers are continuing the work of trying to figure out an explanation that better fits and better predicts human behaviour. Yes, even evolutionary psychology can be self-correcting.