This graph shows the average of 18 non-tree ring proxies of temperature from 12 locations around the Northern Hemisphere, published by Craig Loehle in 2007, and later revised in 2008. It clearly shows that natural climate variability happens, and these proxies coincide with known events in human history.
Loehle also published in 2008 a paper that described why tree rings can not be trusted as a proxy for past temperature variations. Tree ring data have what is called a “divergence problem” in the late 20th Century where the tree ring data data suggests cooling, when in fact there has been warming. This, by itself, should cast serious doubt on whether tree ring reconstructions (such as Michael Mann’s famous “hockey stick” curve) can be used to estimate past global temperature variability.