SPACKLED 2/20/07 Ramsey et al. 2007
Ramsey, G., Bastian, M.L. & van Schaik, C. (2007). Animal innovation defined and operationalized. Behavioural and Brain Sciences, 30, 393-437.
Abstract: Innovation is a key component of most definitions of culture and intelligence. Additionally, innovations may affect a species’ ecology and evolution. Nonetheless, conceptual and empirical work on innovation has only recently begun. In particular, largely because the existing operational definition (first occurrence in a population) requires long-term studies of populations, there has been no systematic study of innovation in wild animals. To facilitate such study, we have produced a new definition of innovation: Innovation is the process that generates in an individual a novel learned behavior that is not simply a consequence of social learning or environmental induction. Using this definition, we propose a new operational approach for distinguishing innovations in the field. The operational criteria employ information from the following sources: (1) the behavior’s geographic and local prevalence and individual frequency; (2) properties of the behavior, such as the social role of the behavior, the context in which the behavior is exhibited, and its similarity to other behaviors; (3) changes in the occurrence of the behavior over time; and (4) knowledge of spontaneous or experimentally induced behavior in captivity. These criteria do not require long-term studies at a single site, but information from multiple populations of a species will generally be needed. These criteria are systematized into a dichotomous key that can be used to assess whether a behavior observed in the field is likely to be an innovation.
Points of discussion from the meeting:
- What is the distinction, for the purposes of operationally defining innovation, between ‘learning’ and ‘memory’? In other words, is learning involved in innovation as the authors suggest, or is it more accurately understood as remembering?
- What role does imagination play in innovation?
- How can disadvantageous and/or maladaptive innovations be identified? Since innovations are not necessarily advantageous, it may be important to know more about the rate at which novel behaviors (innovations?) are ‘thrown out’. For example, due to social pressures against them, as in the case of male cowbirds who are attacked by other males if they sing songs that are really potent (though these songs are better characterized as improvisation or imitation of another’s male’s song).
- This also raises the issue of the importance/role of the social context in the production of innovative behaviors.
- What is the importance of the relationship between imitation and innovation?
I am trying to think of how the operationalisation might help in understanding innovation. Operationalisation is helpful when quantifying something, but I think in most ways of approaching innovation, one would not want to quantify it (Ramsey, et al., 2007, do give exceptions).
If it is to help in talking about innovation, I wonder how delimiting boundaries on what to so label helps, whether it excludes or includes a given behaviour. Surely, Ramsey, et al.’s formula for calling a behaviour innovative or not would not change the way we think about or study the behaviour (though it may remind of some relevant considerations).
I suspect our eagerness to accept Ramsey et al.’s call to define has more to do with our wanting vindication for our arguments for or against animal intelligence, and this may be more distracting than helpful.
Katie suggested I post an email that I sent her earlier (that may or may not be of interest…)
We just submitted a paper on how different song learning strategies (imitation and improvisation-we can’t study invention out in the aviaries but innovation is close to improvisation depending on your definition) are influenced by the social ecology. For example, juvenile males housed with adult female improvise compared to juvenile males housed with juvenile females who copy each other’s song. If you switch the males (so males housed with juvenile females are now with adult females), the juvenile males now housed with adult females start to improvise! We believe this is related to the social behavior of the females as adult females are highly discriminate and do not interact with juvenile males compared to juvenile females who interact at a very high rate with the males. We think the ability to improvise relates directly to the development of song usage. For examples, males who improvise more sing more directed songs to females, an important component in reproduction. I must tell you though the improvisation is only in the fall as the songs become stereotyped in the spring. We do not know the function of improvisation yet-but we have done a few other studies trying to tie the early song learning strategies to the reproductive success. Our hypothesis is that males who learn to improvise early on will have better song capabilities in the spring hopefully leading to better reproductive success.
Im not sure if you talked about vocal innovation in infants (see John Locke comm. after article) but now I am running a study on how the social ecology can influence vocal exploration in prelinguistic infants that is similar to the study described in cowbirds. THe great thing about infants is that they can produce a huge range of sounds!! So for example, if caregivers give non-vocal feedback to every vocalization versus to one out of every 5 vocalizations, how will this influence vocal production? My hypothesis is infants who receive feedback to one out of every 5 vocalizations will explore their vocal range more than when receiving contingent stimulation to every vocalization. In my n=1 analyzed, it looks like this is the pattern-responses to every vocalization results in the infant producing only one type of vocalization compared to every 5th vocalization results in the infant expanding its vocal range.
All in all, we think the infants and birds are trying to get the attention/feedback of
their social partner(s) and produce different sounds for social feedback…
*Side note-Improvisation in the song learning lit. is defined as one individual in a flock producing a vocalization that no other bird is producing.
Great stuff on the cowbirds & infants, Jennifer. If all 5 different vocalizations get a reinforcing response, can you predict which vocalization will become “locked in”? I’m wondering if there’s something like overshadowing going on here. Also, I wonder what would happen if you set up an contingent reinforcement schedule where vocalization A gets the response 100% of the time, B 80% of the time, C 60%, etc.
If all 5 different vocalizations get a reinforcing response, can you predict which vocalization will become “locked in”?
**Not yet…there are a lot of interesting paths you could take on this question. Would all 5 vocalizations get a response from the same individual (as is the case with the infant) or would they be from different individuals (as might be the case with the birds). This really gets down to the question: how does vocal structure and vocal usage relate in the network?
**Also, I wonder what would happen if you set up an contingent reinforcement schedule where vocalization A gets the response 100% of the time, B 80% of the time, C 60%, etc.
***Next infant project…but in songbirds maybe see Burt and Beecher’s work with the virtual social tutor, I think they may have something like this…??? It would be difficult to do this in real time with the birds but much easier with a trained ear in infants.