Do a 1/10 and 1/100 dilution of your sample, then re-run the Nanodrop/spectro. These values should all be roughly the same when multiplied by dilution factor. If they are (which I assume it won't be), take the 1/10 and 1/100 values because these should be similar once multiplied by dilution factor, and are therefore more appropriate for further analysis; you need an appropriate amount of RNA to get correct readings on other wavelengths. This might normalize the 260/280, but I'd also run it on a gel, and confirm that it's RNA. If the well is all lit up, then it's most likely filled with protein as protein has trouble migrating through the gel matrix. This means your prep is actually contaminated, and you should redo the prep, making sure the lysis step was done correctly, and that you don't get greedy during phenol/chloroform, making sure not to go past the top phase. Leave some if you're not sure.
Something is not right. Pure RNA typically produces a 260/280 ratio of 2.1-2.2. Protein contamination increases 280 reading and thus decreases this ration. Since that's not the case, something else is wrong, perhaps estimation. Try re-estimating by using accurate elution buffer for blank. If the problem persists, re-extract.