I'm confused about the details of how the concepts of "putative marriage", "valid marriage", and "invalid marriage" are related in the context of Roman Catholic canon law in medieval times vs. today.
My understanding is that today, a declaration of annulment in the context of the Roman Catholic church is equivalent to an assertion that no valid marriage ever existed between the two parties involved. Here is a description from the website of the United States Conference of Catholic Bishops:
What is an annulment? [...] a Church tribunal (a Catholic Church court) declares that a marriage thought to be valid according to Church law actually fell short of at least one of the essential elements required for a binding union. [...] Why does the Catholic Church require every divorced person who wishes to marry in the Catholic Church, or who wishes to become Catholic, to obtain a declaration of nullity? [...] The tribunal process seeks to determine if something essential was missing at the moment of consent, that is, the time of the wedding. If so, the Church can declare that a valid marriage was never actually brought about on the wedding day.
However, the answer by sempaiscuba to "In medieval Europe, were children born in an annulled marriage automatically illegitimate?" pointed to a quote from Harold J. Berman that contradicts what I thought (I've added a bit more of the surrounding quote that I find relevant to my question):
Where the parties married in good faith, without knowledge of an impediment, the canonists held that the children of the marriage were legitimate and that the marriage itself was valid up to the day it was declared null; this was called a putative marriage.
Berman, Harold J: Law and Revolution, the Formation of the Western Legal Tradition, Harvard University Press, 2009
(Please note, I'm not concerned here with the topic of the legitimacy of the children in such a situation, as modern sources that I've found agree that these children are not considered illegitimate.)
Has there been a change in canonists' definition of annulment, and it was previously understood that a putative marriage was in fact valid up until it was annulled? Or is it possible that Berman slightly misworded his statement here, and putative marriages were not considered in retrospect to have ever actually been valid, but only to have been entitled to the presumption of validity until determined to be null? I would appreciate anyone who can give a pointer to historical primary sources, or to secondary sources other than Berman that address this topic.
When I mentioned my confusion in the comments, sempaiscuba presented an interesting argument to support Berman's wording:
In 1149 Pope Eugene III issued a bull confirming Eleanor of Aquitaine's] marriage to Louis VII was valid. In 1152 it was annulled (with the approval of Pope Eugene) on the grounds of consanguinity.
However, I don't think this answers my question because it seems possible to interpret this as meaning that at the time of the bull, Eugene III presumed the marriage to be valid, and then later on he considered that it had never been valid.