A very insightful thread at WebmasterWorld discusses how Google might distinguish trust and authority notions and how they might influence the algorithm.
In a very summarized manner, the discussion resolves into the following:
|Is based on…
|A hand picked "seed list" of trusted domains" (see also "Trustrank")
|Overall quantity and quality of inbound external links
|Role in the algorithm
|Protects you from "Google-bowling" campaign against your domain
|Matters for rankings
|Query- and topic-dependent factor
This theory can be interpreted in several ways. We can differentiate between trust and relevance where trust is attributed to the domain whereas relevance comes from having backlinks from sites that have similar content who also have backlinks from relevant sites.
Other useful notions to consider are hubs and authority sites: hubs link out to a lot of authoritative documents and authority sites are linked to by a lot of authoritative documents.