We've never been there. Absurdly speaking, there could be a big hologram just outside the Oort cloud, and we wouldn't know. More realistically, in deep space light could propagate by different rules than it does here. Therefore our conclusions are based on chains of hypotheses, only some of which we can test.

True, the alternate options seem so absurd that no reasonable person would argue for them. (Which is exactly why I used that example.) However our current theory requires us to believe that the rules we've measured in our incredibly small corner of the universe work in a much larger volume. Who knows [link|http://en.wikipedia.org/wiki/Pioneer_anomaly|what] unpredicted effects are out there?

Cheers,
Ben

PS Actual parallax calculations are more complex than that paragraph indicates. The biggest confounding factor is stellar aberration, which happens because the motion of the Earth causes light to appear to be coming from an angle. In the same way that raindrops falling straight down do not appear to be falling straight down if you are in a car.