Assuming this is "New England Marriages Prior to 1700" by Clarence Almon Torrey, there is a rather good guide to that 12-volume set at americanancestors.org here (TorreyIntro.pdf).
It's worth reading the entire document as it explains all the quirks of Torrey's indexing system. It includes a number of relevant notes (edited for brevity, bolding added by me):
The bride's name may be presented in a number of ways...
When the bride's full
name is known, without brackets or punctuation preceding or following her
When the bride's surname is within brackets, Torrey had found a reliable source
that identified the bride's maiden surname ...
When the bride's surname is in brackets and preceded or followed by a question
mark, Torrey was less than certain of the bride's identity
Subsequently, we read:
a man or woman might have had more than one spouse in his
or her lifetime, and Torrey presented such cases in a stylized but consistent
manner. A rather extreme example: "Thomas COLEMAN (-1682) & Margaret/
Margery (FOWLER) (OSGOOD) [ROWELL], w Christopher, w Thomas, m/4
Thomas OSBORN 1682; b 1670; Salisbury/Nantucket." In this example,
parentheses were used to show both the woman's maiden name (Fowler) and the
surname of her first husband (Osgood). This entry should be interpreted as
follows: Thomas Coleman of Salisbury and Nantucket, who died in 1682, married
(probably in Salisbury) by 1670, Margaret or Margery Rowell (of whose name
Torrey was certain despite the lack of a marriage record). Margaret or Margery's
maiden name was Fowler, and she married firstly, Christopher Osgood; secondly,
Thomas Rowell, and fourthly, in 1682, Thomas Osborn.
And in the glossary:
w = (previously) wife (and generally widow of)
Based on those notes, and the example, I would conclude that Susannah's previous husband was John Harding, but her maiden name is not known. She was probably (but not certainly) John's widow.
(For clarity: Although the first excerpt declares that brackets indicate confidence in the bride's maiden surname, the example indicates that Torrey used brackets for the surname at marriage and parentheses to show earlier surnames. So I think the absence of a parenthetical surname means Susannah's maiden name was unknown but her surname at marriage to William Brown could be confidently given as Harding, hence that must have been her name from her first marriage to John.)