Hi Herve,

Innate behaviour refers to instinct, a concept widely acknowledged by classic ethologists, such as Niko Tinbergen and Konrad Lorenz. Let me suggest the reading of The Study of Instinct, written in 1951 by Tinbergen.

Any official definition? Well... I do not know whether official, but all behaviours that are basically determined by genes and not influenced by environment and social and learning experiences could be considered innate. This web site is very clarifying: http://www.ck12.org/biology/Innate-Behavior-of-Animals/lesson/Innate-Behavior-of-Animals/

Cheers

Dear Jorge,

Thank you for pointing out Timbergen's The study of instinct. I was lucky to find Chapter 2 here: http://jackknife.med.yale.edu/nsci590-2012/pdfs/tinbergen-instinct.pdf

Tinbergen suggested a method to distinguish innate from conditioned behaviours in page 52, but he didn't provide any committal definition. I note he rejected the notion of interspecies universality (last paragraph "In older text books..."). 

This healthy read made me realise that most of words (provided by the two sources mentioned earlier on) are consequences of the scientific method, and the rest paraphrases or conflicts Tinbergen (and Lorenz):

- Innate behaviour can be observed identically in animals raised in isolation or in the wild. That's where "intrinsic" and "not learned from others / unlearned(!?)" comes from, and from there comes the notion that the trait is "inherited".

- The requirement for repeatability of the experiment naturally leads to "stereotypical, inflexible, and consummate" or "invariant" stimulus and reactions pairs. Consummate stresses that the behaviour is neither learned from other organisms (e.g. by observation) or through iterative attempts when raised in isolation. "Stereotypical" may be "performed in the same way each time by each individual" however significant timing variations are common for a same individual placed in all-things-equal experimental conditions.

- I feel that "Inflexible" conflcts Tinbergen's perspective who points out that some innate behaviours reveal themselves with the maturation of the animal (esp. sexual behaviours).

I note that in "Ambivalent behaviour" (page 50) Tinbergen considered the possibility that any non-null amount of sensory overlap between two distinct releasers US1, US2 for the innate reactions UR1 and UR2 respectively, may cause non stereotypical response when encountered simultaneously. This calls for non trivial quantitative analysis, especially when the number of US->UR pairs increases.

I won't be able to get access to the full book before a few weeks. Did Tinbergen actually form a robust definition of innate behaviour by the end of his book? 

What do you think of innate behaviours defined as being either reflex, taxis, kinesis, fixed action pattern or reaction chain (as in Mazur's Learning and Behavior)? At that point, I am rather tempted to adopt this approach.

Regards

Dear Herve,

I honestly feel that when you have the chance to read the whole Tinbergen's book many of your questions and doubts will be, at least, partly solved. The version I have in my hands is a Spanish translation published in 1970 by Siglo XXI Editores (Mexico). In the first chapter, Tinbergen defines innate behaviour as "a behaviour that has not been mediated by learning processes", and later in the book the author shows how to recognize it (see, e.g., heading "Innate or conditioned?" in Chapter II, a section that you have already mentioned).

I cannot find a robust definition of innate behaviour in Tinbergen's work, apart from the one mentioned above (perhaps, not that robust...). No doubt, you can get quite a few clues of how to define more in detail innate behaviour after reading Tinbergen's book. In this context, it is worth mentioning here Chapter VI, "The development of behaviour in the individual", where Tinbergen elegantly relates innate behaviour to learned behaviour, and notes how innate behaviours that appears in mature individuals (such as reproductive behaviour) may do so after learned behaviours have been acquired. Also, it is very interesting to highlight the predisposition to learn, which is encoded in individual's innate behaviours.

As for the classification you mention of innate behaviours by using terms such as reflex, taxis, kinesis, fixed action pattern and reaction chain —although there are also others, like tropism or motivation— that is an interesting approach, if you are interested in this kind of responses.

In Mandal's Textbook of Animal Behaviour you can read the following: "Criteria for instinctual behaviour are to (a) be automatic, (b) be irresistible, (c) occur at some point in development, (d) be triggered by some event in the environment, (e) occur in every member of species, (f) be unmodifiable, and (g) govern behaviour for which organism needs no training." Interesting recapitulation, isn't it?

Finally, this is also a stimulating reading: http://plato.stanford.edu/entries/innate-acquired/

I hope my comments have been useful to address your questions.

All the best

Jorge

Dear Jorge,

Thank you for searching your own library with such energy! This is very much appreciated. I'll keep an eye out for Tinbergen's book, and I'll check whether my closest library has Mandal's.

Your answer seals the deal. A blend of these perspectives will definitely address my need.

Best regards

Herve

Hi Herve,

This is a very interesting subject! I did some digging on my own out of curiosity, and stumbled upon some resources which you may find helpful.

First is the (curated) article from Stanford's online encyclopedia of philosophy, "the distinction between innate and acquired characteristics", which Jorge also cited. It seems to provide good framing of the issue and cites various ethological researchers, etc. http://plato.stanford.edu/entries/innate-acquired/

The second is a 1979 (!!) review paper by John Cassidy, "Half a Century on the Concepts of Innateness and Instinct: Survey, Synthesis and Philosophical Implications", which may prove to be a good resource for work done until then:

http://onlinelibrary.wiley.com/doi/10.1111/j.1439-0310.1979.tb01038.x/abstract

Armed with the knowledge that "innateness" is often used as a keyword for this literature, I did another search and came up with a more recent review, this time by Mameli & Bateson (the latter being a familiar figure in this field), "An evaluation of the concept of innateness" (2011).

http://rstb.royalsocietypublishing.org/content/366/1563/436

Again, it can also be mined for references at the very least.

Another key term is "fixed action pattern" or "FAP", which Cassidy refers to in his review article above. A quick search shows that this term can be used to search for work that specifically examines FAPs or "innate behaviours" in the invertebrate and comparative literature, e.g., with bees and wasps but also mice. 

For instance, this paper looks at instinctive grooming as a FAP in mice that is used as a baseline to compare whether there are obsessive-compulsive behaviours that also exhibit sequencing of the kind seen innately with grooming behaviours:

http://bmcbiol.biomedcentral.com/articles/10.1186/1741-7007-3-4

On this note, the esteemed neuroscientist Rodolfo Llinas wrote a book that discusses evolutionary origins of the central nervous system and prominently features FAPs, and may thus be of use, indirectly, as well:

http://www.amazon.com/I-Vortex-From-Neurons-Self/dp/0262621630/

Depending on where in the evolutionary tree you wish to focus on, you can use these different terms to refine your search.

Best of luck!

My thanks again to Vlad and Jorge for their useful pointers. Reading Tinbergen and Lorenz is enlightening to understand where subsequent legions of researchers are coming from. Peer reviewed papers soon reveal that the study of behaviour was influenced by tectonic forces -- historical events and cultural preferences -- and morphed accordingly.

I concluded the journey by reading Frans de Waal, who remains thought provoking and confirms that no-one can avoid a discussion on behaviour to turn into a debate.

According to RG, this thread has attracted 200+ readers since its creation, that is 2 readers a week. I started it while researching a specific angle for my Ph.D. research, and never stopped really.

My angle -- my bias! -- is animal behaviour, and ethology. Humans being mammals, i.e. animals, all the following should be of interest for people who are interested in behaviour of the human species. We are part of a big continuum.

There are several definitions of "behaviour" and "innate behaviour" to be found from different yet symbiotic fields, e.g. ethology, comparative physiology, psychology, physiology, ecology, genetics. This is one of many key words that are overloaded with too many connotations, that make discussions so difficult between peers working in different silos, and so remarkable in multidisciplinary teams.

Both Nikolaas Tinbergen and Kenneth D. Roeder noted that the word behaviour carried different connotations for the researchers in the fields of ethology, psychology, and physiology.

Conceptually an ethologist observes intact animals in their natural habitat and creates an exhaustive nomenclature for each species' activities: the ethogram. Given a species' ethogram one can record sequentially all the activities performed by an animal. The knowledge of each activity's predecessor allows to perform a simple statistical analysis to identify stereotypical sequences of activities and calculate how likely they may occur. These flowcharts are known as kinematic diagrams. Ethologists categorise activities that are not modified by experience as innate. "Not modified by experience" is indeed a short hand that should be understood as not modified permanently, and not modified in other characteristics than the frequency and intensity of the response. Innate responses may be modified transiently, as seen with habituation and sensitisation.

The innate activities are subsequently studied in order to discover the combinations of external and internal stimuli that trigger each of them. For this reason, activities are also known as responses. An innate behaviour is actually a pair {stimulus, innate activity}, although it is frequent to find the word behaviour used as a synonym for activity (and for response).

Two examples of innate behaviours are provided by the herring gull. The gull will incubate any rounded object the size of an egg that may be in the nest. Also, the gull will drive any red object out of the nest. Ethologists and comparative psychologists consider that all animals share innate behaviour patterns that include reflexes, taxes, kineses, fixed action patterns, and reaction chains. This applies to all animals including those with the least amount of neurons e.g. C. elegans with 302 neurons, or nerve nets like hydra.

The categorisation of {stimulus, innate activity} pairs as behaviours implies that a behaviour is stereotypical. Stereotypical is defined as a sequence of unconditional ordered reaction(s) that follow an unconditional stimulus — the releaser — where the timeframe between the stimulus and the last reaction may vary for the same individual and the same environmental conditions, as long as the causal effect is clear. The element of causality is often referred as purposive. One must note that stereotypical must not be understood with the meaning of deterministic, for behaviours never occur 100% of the time, even for the simplest monosynaptic reflexes like our knee-jerk reflex.

It is interesting to note that behaviours can be ambivalent. Behaviour ambivalence results from the presentation of stimuli that release responses whose purpose conflict with one another. If a red rounded object the size of an egg is planted in the nest, the gull will alternatively drive it out of the nest and incubate it. The nocturnal moth who flies toward females while being hunted by a bat must also experience ambivalence. Ambivalent behaviours are found in all animals, including humans. This is typically what happens when humans fall while holding a baby or a fragile object: we tend to try and perform two actions that are impossible to carry out simultaneously. The first action is an attempt to protect what we carry with our arms, and the second action is an attempt to extend our arms in order to attenuate the impact.

In the case of the gull, the stimuli are external, as is a bat attack for a moth. However the moth's reproductive drive is definitely internal. Hunger is yet another example of internal stimulus. Internal stimuli are endogenous to the nervous system, whether they are triggered by so called pacemaker neurons or through the release of hormones.

It would therefore be a mistake to consider that ethologists see animals as automata, for the acknowledgement of internal stimulation is ground to assert that the nervous system is stateful: there is behavioural evidence — with innate and conditioned behaviours — that the effects of internal and external stimulations last longer than the complete realisation of the stereotypical responses.

Finally, innate behaviours are intrinsic. A behaviour is said to be intrinsic when animals raised in isolation without any means to learn the behaviour from a peer perform the same stereotypical and purposive behavior as animals raised socially. The famous deprivation experiment conducted on the squirrel Sciurus vulgaris L. by Irenaus Eibl-Eibesfeldt offers a dramatic example. The squirrel is known to bury nuts according to the unmistakable fixed action pattern that follows:

"This animal buries nuts in the ground each fall, employing a stereotyped sequence of movements. It picks a nut, climbs down to the ground, and searches for a place at the bottom of a tree trunk or a large boulder. At the base of this object it will scratch a hole by means of alternating movements of the forelimbs and place the nut in it. The nut is rammed into place with rapid thrusts of its snout, covered with dirt by sweeping motions, and tamped down with the forepaws."

Eibl-Eibesfeldt raised squirrels in isolation and observed the following:

"When an animal is raised under these conditions, it is able to complete this behavioral pattern on its initial attempt. If it is given nuts, it eats until satisfied; however additional nuts are not drooped but carried in its mouth seemingly in search of a place to deposit them. Vertical structures frequently provide the squirrel with such an opportunity. When it finds a suitable location, the animal deposits the nut, pushes it into a corner or niche with its snout, and makes the covering and tamping movements with the front legs, although it does not dig a hole."

Like the ethologist, the psychologist studies the whole animal albeit with a focus on cognition through experiments based on learned behaviours. Learned must be understood as the temporary or permanent modification of an innate behaviour, either by modifying the stereotypical activity related to a given stimulus, or making a new stimulus produce an activity stereotypical of another stimulus.

As Tinbergen's contemporary physiologists, most of today's neuroscientists work with tethered animals in order to record neural activity, or with animals pinned and stripped of body parts in order to override the sensory organs or the muscles involved in a narrow set of behaviours. They work on both innate and learned behaviours and outside of the animal's natural environment. Neuroscience also harbours a host of different computational simulations, whose contributions to bridging physiological mechanisms and behaviours are variable:

• attempts to replicate the workings of circuits under normal and disrupted conditions using a variety of neuron models,
• attempts to replicate locomotory patterns with models inspired from biology and engineering,
• attempts to build systems that are both homologous and remote from the living in an attempt to infer from the artificial system the characteristics required from the living to make a decision,
• attempts to comprehend the evolutionary process.

In the words of Frans de Waal "Each organism has its own ecology and lifestyle, its own Umwelt, which dictates what it needs to know in order to make a living.". An Umwelt is both what an animal can sense from the environment and how it can change the environment, according to its own sensory and motor capabilities. Behaviours are necessarily the products of the co-evolution between the environment and the animal's nervous system.

Larry Swanson also advocates this perspective. His "Four systems network model of nervous system organization" diagram provides additional directions: "... behavior is a function of the motor system. (...) The second premise is that there are two major classes of functional inputs to the motor system: sensory or reflex, and cognitive or voluntary.”.

One must note that C. elegans has neurons that fill in several roles: sensory, motor, and inter-neuron. Therefore Swanson's second premise may not be taken too literally.

--

My conclusion is that one has to shape a definition of behaviour in one's own field, for one's specific purpose, and be upfront about the differences that other researchers in other fields may see and resent. On the research plane, we are again part of a big continuum.

Automatic Cognition? Social Psychology term for non-calculated response.