This "Journal of the National Cancer Institute" looks pretty legit (Oxford publication) but the study itself is pretty dumb.
http://jnci.oxfordjournals.org/cgi/content/abstract/djn514v1 - Go here to read the non-media version.
A) It is ridiculously short for a scientific paper.
B) I can't find any data (tables/figures) to support their conclusions. It is pretty sketchy
C) The only thing this study showed is a correlation between alcohol consumption and cancer IN WOMEN. Nothing about men.
D) If you read the journal, breast cancer incidence due to alcohol consumption rose by 11 cancers per 1000 cancers (or women? not sure, wording is awkward) for BREAST cancer
E) All of the other tested cancer incidence rates rose by approximately 1 per 1000. That's a 0.1% higher risk. That's within random error for a study like this. I call biased bullshit.
All this confirms to me is alcohol consumption raises risk of breast cancer in women. Period.