Many researchers conclude this skill results from core innate knowledge, because it is too complex for infants to learn from the beginning.
"Yet we find certain things seem to be very important and kids do seem to be learning," said Professor Ted Ruffman, who will lead the study.
"We've got clear evidence that kids are learning, so we are now just looking to see exactly how they are learning - and we can film the world from the baby's perspective, as whenever the camera is pointing somewhere, about 85 per cent of the time, the baby is looking at something."
Ultimately, Professor Ruffman and two colleagues aim to find out if the social worlds of infants are structured enough to gradually build theory-of-mind, and what information is likely to be most important for them in doing so. After capturing what the infants look at for the first three years, the team will use the data to examine how caregivers build an infant's environment based on repeated behaviours and how they talk about these behaviours.
Performance on theory-of-mind tasks has been linked to increased empathy, peer acceptance and social skills, meaning that this study could potentially have practical implications for caregiving strategies that help promote children's abilities in these areas.
Other highlights of projects of benefit from New Zealand's premier funder of investigator-initiated research include a volcano model, dubbed Mount Doom, that will help scientists simulate devastating pyroclastic flows; the science behind the brain drain, and a study looking at the potential of self-healing, sticky, stretchy electronics.
Marsden Fund Council chairwoman Professor Juliet Gerrard said while the number of researchers' proposals put forward for this year's $53 million pool was slightly down on last year's application rate, the council still reviewed more than 1200 bids from New Zealand's research community.
A subset of 208 proposals progressed to the second round, before the final 92 were selected, making for an overall success rate of 7.7 per cent.
"We look forward to seeing these new projects take shape over the next three years and learning what the researchers discover and how this might benefit New Zealand in the long term," she said.
Outsmelling best sniffers
Could we out-sniff a dog? Plant and Food Research's Dr Colm Carraher aims to create an artificial nose so strong it could rival or even beat canines and insects.
Smell is based on odorant receptors. These receptors chemically bind to odour molecules, triggering a reaction inside the olfactory cell, which generates a voltage which our brains recognise as a smell.
Although dogs are often used to aid our feeble sense of smell, insects are far better - a moth will turn upwind when a single molecule from a desirable source strikes its antenna.
Dr Carraher has worked out a way to produce odorant receptors from fruit flies and place them in artificial membranes.
With a $300,000 Marsden Fast-Start grant, he plans to attach these receptors to sensor surfaces to create an artificial nose.
Designed primarily for horticultural purposes - "smelling" a fruit's ripeness - its applications could range from medical tests to sniffing out biosecurity threats.
