Page 9 of the AR4 SPM says that "Palaeoclimatic information supports the interpretation that the warmth of the last half century is unusual in at least the previous 1,300 years". The relevant chapter of the main WG1 report is chapter 6, "Paleoclimate", and in particular section 6.6, "The last 2,000 years". Although the IPCC does acknowledge some warming in medieval times, Box 6.4 on page 469 says "However, the evidence is not sufficient to support a conclusion that hemispheric mean temperatures were as warm, or the extent of warm regions as expansive, as those in the 20th century as a whole, during any period in medieval times".
But, ironically, the data used by the IPCC in fact supports exactly this conclusion! Here is their Figure 6.10b, showing various 'proxy reconstructions' of temperature in the Northern hemisphere. Notice that these show a clear temperature maximum around 1000 AD, corresponding to the medieval warm period (MWP):
The data from which this figure was produced have been made available on Tim Osborn's website. There are six proxy datasets that cover the MWP: B2000, DWJ2006, ECS2002, HCA2006, MJ2003 and MSH2005. Averaging these six gives a series from 831 AD - 1960, which looks like this:
Note that the series has not been smoothed at all. Smoothing is a Bad Thing, according to statistician William Briggs, and introduces problems of what to do near the end-points. This graph ends in 1960, since the HCA2006 series ends then. If we omit HCA2006 and just average the remaining five datasets, we can go up to 1979, and we get a similar picture:
Thus, the IPCC's own data shows a clear MWP around 1000 AD, and shows that the MWP was as warm as the late 20th century, completely contradicting the statements above from chapter 6 and from the SPM.
So how does the IPCC reach these conclusions that are contradicted by their own data? By employing the IPCC's favourite trick - measure something one way in one period (proxies for the MWP) and compare it with a different method in a different period (direct temperature measurements for the last century), disregarding the fact that these two methods cannot be compared because of the "divergence problem".
See this page for more supporting evidence for the medieval warm period.