One part of the scientific method is concerned with how to sift through research results and theories that are mostly flawed or otherwise wrong in order to extract, further explore, and (where possible) prove the results that have value. Consensus amongst scientists is important to the scientific community because independent replication of results is important to the scientific method. Nothing should be taken at face value, or as more than a possibility, until there is a weight of research from diverse scientific groups to support it.
The scientific community doesn't produce an output of nice, neat tablets of truth, pronouncements come down from the mountain, however. It produces theories that are then backed by varying weights of evidence: a theory with a lot of support stands until deposed by new results. But it's not that neat in practice either. The array of theories presently in the making is a vastly complex and shifting edifice of debate, contradictory research results, and opinion. You might compare the output of the scientific community in this sense with the output of a financial market: a staggeringly varied torrent of data that is confusing and overwhelming to the layperson, but which - when considered in aggregate - more clearly shows the way to someone who has learned to read the ticker tape.
But even consensus itself isn't a clear-cut item. We live in a complex world, and as pointed out over at Depressed Metabolism, some reaches of science can present challenges to the normally safe approach of interpreting the present state of knowledge through scientific consensus:
Scientific consensus seems a reasonable concept. If a great number of individual scientists arrive at a similar opinion this is generally a sufficient reason to have confidence in those views. Skeptics about scientific consensus often use examples of scientific views that started out as a minority view to become the majority view later. Although these examples raise interesting questions about how science evolves as a collective undertaking, they cannot be used to argue against the importance of scientific consensus as such. For every minority view that became a majority view there are a lot more examples of crackpot theories that are still crackpot theories today. Nevertheless, there are a number of situations where the concept of scientific consensus is of limited value.
A good example are fields that are so interdisciplinary that there is no clearly identifiable group of scientists who can be perceived as authorities on the matter. For example, what is the scientific consensus on cryonics? The consensus of biologists? The consensus of cryobiologists? The consensus of neuroscientists? The consensus of experts on nanotechnology? The consensus of those who study cryonics in all its aspects? It is clear that when there is no clearly identifiable group of experts, the concept of scientific consensus becomes problematic.
This interdisciplinary issue of consensus and shared information is becoming ever more an issue in the life sciences, and in particular in aging research. Human biology is so complex that entire academic lifetimes can be - and have been, and are - spent on cataloging very tiny parts of the whole. The cutting edge of biology and biotechnology research is conducted at the very tips of the widespread leaves, that are at the end of the widespread twigs, that are at the end of the widespread branches of biology as a field. But many widely separated twigs are still related in the processes of aging - and that is an issue, because there is too little communication and synthesis taking place between these many disparate enclaves of knowledge. Biology as a field needs, and is in fact entering, an era of synthesis, in which drawing together existing knowledge is just as important to progress as generating new knowledge:
Scientific progress goes through cyclic periods of fragmentation followed by synthesis. During fragmentation, many different groups toil away at their own little pieces of the great unknown. Each group generates data that, at first, appears to have little to do with other efforts. As the data piles higher, correlations start to appear - and so do the efforts at synthesis. Gradually, the focus in a field shifts from finding new information to making sense of what is known, pulling it all together such that links, correlations, and chained mechanisms are understood. Then they next great unknown beckons and the process of fragmentation starts once more.
At present the grand study of human biochemistry is moving from fragmentation to synthesis. It is still the case that some different specialties know little of one another's work. Researchers look at the same mechanisms and compounds, giving them different names and assuming different dominant roles in biochemical processes, all the while missing out on the enlightenment that a complete picture can bring.