While various studies seek to quantify the anthropogenic influence on extreme precipitation in the historical record, there is a need for improved understanding of the (at times) conflicting conclusions on the presence of human-induced climate change on extreme precipitation. Here, we introduce a single framework for characterizing the historical signal (anthropogenic forcing) and noise (natural variability) in seasonal extreme precipitation over the last 120 years. An important aspect of our analysis is that we simultaneously isolate the individual effects of seven modes of climate variability while explicitly controlling for joint inter-mode relationships; furthermore, we utilize a spatial statistical analysis to translate trends from stations to a high-resolution grid. We find that (consistent with related work) the frequency of seasonal extreme precipitation is generally increasing for CONUS overall, but there are important local deviations from this trend. At a high resolution, we are able to isolate and detect a human influence on seasonal extremes in DJF and SON in spite of extremely large background variability. Finally, we use our framework to explain conflicting trends in extreme precipitation depending on the time period chosen for calculating trends.