Having a genuine 'Dark Age' is actually quite rare; for most of post-Roman Europe we know what's going on somewhere most of the time, even if we don't know what's going on everywhere all of the time. In Francia, for example, Gregory of Tours chronicles much of the sixth century that is 'Dark' in England, while the Byzantine history of the period is chronicled in depth by writers such as Procopius and Agathias, who (especially in the case of Procopius) also do their best to explore their wider European context.

The frequently given example of an actual 'Dark Age' is in Early Medieval Britain; the roughly 60-70 years between Gildas' De Excidio Britanniae and the Augustinian mission of 597 when Procopius is our other only faint and guttering candle in the darkness. This doesn't mean that we make things up though.

One of the biggest problems in early medieval history is that of survivorship bias: a tendency to think that anything that survived this long must have been important, or, more perniciously, that the lack of survival means something wasn't important, or indeed didn't exist. One example is law codes: starting with Æthelberht of Kent's post-conversion codex of c.597, we have a reasonably substantive corpus of Anglo-Saxon legal codes from the sixth to the eleventh century, but it is far from complete. We have, for example, no extant copies of the laws of Offa of Mercia, one of the most important Anglo-Saxon rulers and perhaps the defining political figure of 8th century England. Indeed, we only know that Offa had a codex because Asser's Vita Ælfredi tells us that, in preparing the West Saxon legal reforms in his Doomboc, King Alfred studied extensively from Offa's legal codes and drew heavily from them. Clearly, in a 9th century context, the laws of Offa were a prominent and influential document, but to a fleeting glimpse from modern eyes, 8th century Mercian law is a 'Dark Age'. We can, of course, use some of Alfred's Doomboc and earlier Mercian laws to extrapolate what Offa's legal codex might have looked like. Similarly, we can use Æthelberht's 597 codex to infer the basic tenets and structure of pre-literate Kentish (and wider Anglo-Saxon) laws and customs, even though we have no extant sources.

Why is this important? Because just because we don't have contemporary sources for those periods now doesn't mean that they didn't exist at the time. While there was little central organisational hierarchy in the Insular Christianity that survived in Britain until the Augustinian mission of 597, the legacy of Roman conversion meant that some literary sources must have survived: The Welsh Chronicle Annales Cambriae focuses (as might be expected) on Wales, but also periodically offers glimpses into England from the early 6th century onwards. In Ireland and Scotland, the Vita Columbae was written long after that saint's death, but is believed to be based on earlier non-surviving chronicles of the period. The Anglo-Saxon Chronicle draws on Bede's Historia Ecclesiastica for much of its early history, who in turn draws on Gildas' de Excidio, but both contain enough differences and additional 'flavour' that we have to infer they also drew on other surviving histories at the time. Similarly, later eleventh and twelfth century histories contain additional details that either hint at great creative liberty or a wider corpus of source texts than we enjoy today.

Our other window on the period is, of course, archaeology. To pluck one random example, excavations at Eddisbury hillfort suggest that there was some degree of Irish occupation or settlement of the Wirral and the Cheshire Plain following the abandonment of the legionary fortress at Chester. Gradual changes in grave goods and material culture can help us estimate the scale and pace of Anglo-Saxon migration and settlement, and the process of Anglicisation. We know that Roman cities like Wroxeter were largely but not entirely abandoned, that some districts were occupied and fortified internally, and that there was periodic reoccupation of Iron Age hillforts.