To search for a word in a dictionary without a proper index is like looking for an address in a city without a decent map. Abstract Dictionaries are repositories of knowledge concerning words. While 1 readers are mostly concerned with meanings, writers are generally more concerned 2 with word forms expressing meanings (lemma). I will focus here on this latter task: 3 building a tool to help authors to find the word they are looking for, word they may 4 know but whose form is eluding them. Put differently, my goal is to build a resource 5 helping authors to overcome the Tip-of-the-Tongue problem (ToT). Obviously, in 6 order to access a word, it must be stored somewhere (brain, resource). Yet this is 7 far from sufficient. Access may depend on many other factors than storage of word 8 forms: organization of the dictionary (index), the user's cognitive state, i.e. available 9 knowledge at the onset of search, the distance between the source-and the target-10 word (direct neighbor or not) , the knowledge of the relationship between the two, 11 etc. I will try to provide evidence for the claim that (word) storage does not guar-12 antee access. To this end I will compare a well-known lexical resource, WordNet 13 (WN), to an equivalent one, but bootstrapped from Wikipedia (WiPe). While both 14 versions contain basically the same set of words, the latter contains many more 15 (syntagmatic) links than WN. This is probably the reason why WiPe outperforms 16 WN. In the last two sections I will explain under what conditions WN is suitable 17 for word access, and what it might take to go beyond the limitations of this famous 18 resource.