Is learning a characteristic of all organisms?

[1MileCrash]

Just heard someone make that claim...seems a little bit, well, stupid.

Do all organisms possesses the ability to learn new things? I would argue that moss on a rock, any plant (note: naturally responding to environmental change is not learning) etc. do not learn..

 

[apeiron]

Just heard someone make that claim...seems a little bit, well, stupid.

Do all organisms possesses the ability to learn new things? I would argue that moss on a rock, any plant (note: naturally responding to environmental change is not learning) etc. do not learn..

It of course depends whether you define learning narrowly or broadly. A general mid-range definition would be "modification of behaviour due to experience".

If you are looking for a sharp dividing line, then not having a central nervous system would seem to rule out the plant kingdom.

Except plants may then have some level of integrative memory in the form of signal-transduction networks...

Signal-transduction networks share properties with neural networks, and the learning parallels can be drawn easily (22). Neural networks learn by increasing the numbers of connections (and the strength of the connections) between the neurons representing the chosen path to connect signal and response. The result of learning (reinforcement) is to accelerate the information flux rates between the signal and the response. Elevating Ca21 transduction constituents is analogous to increasing the numbers of connections between neurons in a neural network. The increased information flow that results represents a kind of cellular learning. This cellular learning, coupled with the memory built into signal transduction systems (22), suggests an unexpected form of cellular intelligence (23).

http://www.pnas.org/content/96/8/4216.full.pdf

 

[atyy]

Phenotypic plasticity
http://jeb.biologists.org/content/209/12/i.full

There are different sorts of learning like associative/non-associative etc and different sorts of memory like declarative/procedural etc. so you'd probably like to refine your question according to one of those taxonomies. apeiron's turf, I believe, since I think he was a neural networks guy.

 

[mishrashubham]

Building upon what apeiron and atyy said, I would say that a basic principle in cellular signalling is that signalling pathways have evolved to respond to changes in stimuli rather than their absolute level. This acts like an effective filtering mechanism you could say, which gives accurate responses to stimuli and ignores background noise. Every time a receptor is activated it needs to be reset in order to accept new signals. Repeated cycles can cause them to desensitize over time. Cells usually accomplish this by blocking out receptors by special molecules or phosphorylation, for example repeated stimulation of β-androgenic receptor causes it to bind with β-Arrestin and prevents it from binding to hormones. This might very well be called molecular or cellular learning.

You can even extend this to the immune system with their memory B cells and although that wouldn't be considered an organism.