Australasian Science: Australia's authority on science since 1938

Do Fish Feel Pain?

Illustrator: Marsha Wajer

Illustrator: Marsha Wajer

By Brian Key

If you want to know whether your pet goldfish can feel pain you had better look inside its head to see if it has the brains for it.

An American folktale tells of how the Good Lord created fish and filled the rivers with these creatures. The finest and most handsome fish of all was the catfish. The Good Lord made catfish with brilliantly coloured scales that sparkled hues of red, blue and gold as they swam in the shallow waters. The Good Lord enjoyed walking along riverbanks and seeing these beautiful creatures. Then one day the Good Lord came upon the Devil who was seated by a river. The Devil was plucking the scales off a catfish so he could fry it for dinner. The catfish was wincing with pain as each scale was pulled out and this angered the Good Lord. The Good Lord threw the catfish back into the water. To ease its pain, the catfish swam to the bottom of the river and slithered about in the mud. The pain disappeared and from that day on catfish no longer had scales and instead became covered in slime, making them difficult for the Devil to catch.

It’s very easy to conclude that a fish feels pain if it winces every time a scale is plucked. Plucking an eyebrow or pulling sticky plaster off your skin causes a twinge of pain, so it’s not unreasonable to assume that a grimacing fish also feels pain. However, looks can be deceiving. The question of whether fish feel pain has become a heated academic debate involving scientists and philosophers from diverse fields including neuroscience, psychology, medicine, ethology, systems biology, animal welfare and robotics. The resolution of this question has important implications for understanding the nature of consciousness.

Given that fish respond to stimuli that in humans cause pain, why is the question that fish feel pain such a hard problem? Assessing pain in humans seems to be a relatively simple matter. All we need to do is to prick someone’s big toe with a pin and ask whether they felt pain.

But what happens if that person can’t speak and tell us what they are feeling? Again that seems to be easily resolved. We could watch the person to see if they withdrew their foot following the pinprick. Measuring the distance moved by the foot could provide a quantitative index of the amount of pain. But this is where it actually begins to get tricky.

What happens if the person has suffered a spinal cord injury and they can’t feel any sensations arising from their feet. Some paraplegics with complete spinal cord injury can still physically respond to stimuli, such as a pinprick to the sole of the foot, despite not experiencing any pain sensation. These bodily responses occur non-consciously (i.e. without being aware or feeling).

Many different types of human behaviours can occur non-consciously. Extreme examples include walking and driving while asleep. A sleepwalker can get out of bed and walk around objects and even operate machinery while being totally oblivious to what they are doing. Observing body movements is clearly not an appropriate way of assessing the feeling of pain.

So how do we interpret the response of the catfish to the Devil’s plucking of scales? We have learned that merely observing behaviours can be misleading. A better way of inferring pain is needed.

In 2003, Lynne Sneddon in the United Kingdom thought she had found this when she began observing the response of fish to injections of acetic acid (vinegar) into their lips. She injected five trout and found that they subsequently displayed anomalous behaviours not seen in fish injected with saline (salt water). The acid-injected trout appeared to rock on their fins or rub their lips into the gravel on the bottom of the holding tank, much like the catfish in the folktale. Although they only performed these movements less than two times per minute, she still likened it to the way humans sometimes vigorously rub an injured area. She then found that these behaviours seemed to disappear if the fish were subsequently injected with morphine. This study is often touted as definitive evidence that fish feel pain.

Corollaries between fish and human behaviour are easily imagined. Given what we know about the problems of merely observing behaviours, we need to be very cautious in interpreting preliminary observations of a small number of trout.

In a lesser-known and subsequent study undertaken in 2008, Sneddon had to increase the concentration of acetic acid 50 times before she observed the occasional anomalous behaviour in carp (and this time only in two of the five fish injected with acid). In contrast, zebrafish never showed any anomalous behaviour after acid injection.

In her latest repeats of this experiment in 2011 and 2014, Sneddon failed to report this anomalous behaviour in either trout or cod after acid injection. These disparate results highlight the inherent difficulties in drawing conclusions too early from casual observations of a small number of fish.

We also need to be circumspect when extrapolating from the effects of analgesics since they can act in a variety of ways. They are known to disrupt the early stages of transmission of signals from the skin to the brain (which is analogous to how cutting the cord that connects a keyboard to a computer prevents signals from activating central programs). Altered behaviour following drug administration merely informs us that there has been a break in the transmission of information somewhere within the nervous system. It does not indicate that fish feel pain. While drugs may prevent a paraplegic with a complete spinal cord injury from responding to a pinprick, it is clearly wrong to subsequently conclude that they can feel pain.

If we can’t trust behavioural observations and we can’t trust the use of analgesics, then what are we expected to trust? To answer this question we need to clarify where in the brain the feeling of pain emerges. Modern human brain imaging studies have conclusively revealed that pain is dependent on highly dynamic and complex patterns of electrical activity within the forebrain. These electrical signals travel along an intricate network of fibres that precisely interconnects various specialised forebrain subregions.

In some respects, the human forebrain is like a major metropolitan road network. During the day (akin to being conscious and aware) there is traffic flowing back and forth along a grid consisting of motorways between large districts, major arterial roads between suburbs, and side roads within the suburbs. In contrast, at night (comparable to being non-conscious and unaware of pain) traffic is reduced, flows mostly in one direction, and is predominantly restricted to major roads.

The answer to the question about what we can trust as a measure of the ability of fish to feel pain now turns out to be quite simple – it has its origins in the study of anatomy. We can ask the question: are fish equipped with the specialised forebrain subregions as well as the bidirectional network of fibres and interconnections that are necessary for awareness and the feeling of pain?

Before we continue any further, a preparatory lesson is needed on the anatomy of a generic vertebrate brain. For our purposes, the brain can be divided into a front part called the forebrain and a rear part called the brainstem (Fig. 1). The brainstem is the evolutionary old part of the brain and is present in both fish and mammals. Given that fish and mammals shared a common ancestor about 450 million years ago, its preservation over this timescale highlights the functional importance of this brain region. The reason that the brainstem is so vital is that it controls most of the basic life-sustaining functions, such as respiration, heart rate and locomotion, as well as sensory reflexes associated with vision and hearing.

But things are vastly different in the forebrain, which is so disparate in fish and humans that it is difficult to find any structural likenesses (Fig. 1). Given that their anatomies are so dissimilar it is not surprising that their functions are also different. This is best exemplified by experimental reports of goldfish that have had most of their forebrains surgically removed. The behaviour of these animals cannot be easily distinguished from fish with a functional forebrain: they eat normally, swim normally and respond to danger normally. They are just as hard to catch with a fish net in a tank as normal goldfish.

In contrast, lobotomised people have severely altered behaviours. The movie One Flew Over the Cuckoo’s Nest highlighted the detachment and loss of spontaneity observed in psychiatric patients following lobotomy, which involves severing discrete connections in the forebrain.

As already mentioned, complex connections between subregions of the forebrain are responsible for generating the feeling of pain. These circuits can be artificially stimulated with surgically implanted electrodes, causing people to feel pain. In contrast, when these circuits are damaged by ischaemic stroke (a blocked artery in the forebrain), the sensation of pain can be lost.

Pain circuits similar to those in humans are found in other mammals, from mice to monkeys. However, they are not present in the fish forebrain and can’t be found in other parts of the fish nervous system. Indeed, no other brain circuits have been identified in fish that could take over the function of these unique pain circuits.

It may be helpful to think of another computer analogy. Low-end desktop computers lack the processing power to solve difficult computations because they contain microprocessors with relatively simple circuits. The only way to perform such computations is to upgrade to a more expensive computer with advanced microprocessors that possess more complex circuits.

The ancient science of anatomy has enabled us to look deep inside the heads of fish and humans and to realise that the fish brain lacks the necessary hardware for feeling pain. The myth that fish feel pain has been busted. The catfish’s wincing and slithering behaviour, in response to the Devil’s plucking of scales, was performed non-consciously using ancient survival circuitry in the brainstem.

While some people will continue to be disbelievers, the history of science reminds us that progress is achieved when we abandon intuitive beliefs about the way we would like nature to be, and instead accept nature for what it is.

Professor Brian Key is Head of the Brain Growth and Regeneration Laboratory in The University of Queensland’s School of Biomedical Sciences.