Lisp
A lisp is a speech impediment. People with a lisp are unable to pronounce sibilants (like the sound of "s") and replace them with interdentals (like the sound of "th"). It is somewhat ironic that this handicap is called a "lisp", as this is unpronounceable to those with the condition.
A lisp in Spanish?
There is a widespread misconception, especially among Latin American speakers of Spanish, that speakers of Castilian Spanish "speak with a lisp". According to this urban legend, which you will not hear from a Spaniard, the lisp became common in Castilian because one of the Spanish kings (generally identified as Felipe V or Carlos V) spoke with a lisp, and his courtiers did not want to embarrass him by speaking otherwise. Actually, if speakers of Castilian Spanish spoke with a lisp, they would be unable to pronounce the sound [s] ("s") and would substitute a [θ] ("th") for it, but reality is that they pronounce and distinguish those two sounds just as fine and sharply as English speakers, and if a [θ] is pronounced in place of a [s], the result is a completely different word in Castilian, as reflected in the orthography (caza ≠ casa), just like what happens in English (thing ≠ sing). If speakers of Castilian Spanish were speaking with a lisp just because they pronounce the sound [θ] (and distinguish it from [s]), then speakers of standard English would also be speaking with a lisp because they do just the same.
The real story of why Castilian speakers distinguish those two sounds while Latin American and Andalusian speakers do not is quite different and has nothing to do with lisping. 15th-century Spanish had six sibilant phonemes, more than any current variety of Spanish, and those six phonemes merged differently as they evolved into the pronunciation of the modern dialects. There were three pairs of voiceless versus voiced sibilants: dentoalveolar affricates (spelled c/ç vs. z), apicoalveolar fricatives (-ss-/s-/-s vs. -s-), and prepalatal fricatives (x vs. j/g). The first step away from that system was to fricativize the dentoalveolar affricates.
Then, in Castilian Spanish the second step was to lose the voiceless/voiced distinction in favour of the voiceless member, and the final step was to alter the pronunciation of the three resulting phonemes so as to boost their acoustic distinction (because they were used to distinguish many minimal pairs, but phonetically they were too similar and thus prone to cause confusion). The dentoalveolar was moved "forward" to interdental (the sound of th) and the prepalatal was moved "backwards" to velar (the sound of ch in Scottish loch), resulting in the three-way distinction used in modern Castilian: interdental /θ/ (spelled c/z), apicoalveolar /s/ (s, note that this sound of Castilian is different from an English s, because it is pronounced with the tip of the tongue instead of with its blade), and velar /x/ (j/g).
In Andalusian, however, the phonological evolution since the fricativization of the affricates differed from that of Castilian. The second step in Andalusian was to merge the members of the dentoalveolar and apicoalveolar pairs according to voicing, before voicing was lost (the voiceless c/ç and -ss-/s-/-s merged together, and the voiced z and -s- merged together), which resulted in only one merger phoneme /s/ (instead of the two, /θ/ and /s/, that resulted in Castilian) for all of the four original sibilants once the third step (losing voicing) took place. The single phoneme /s/ that resulted from all this merging was pronounced differently in different parts of Andalusia; in some places (areas of seseo), it sounds just like an English s, and this was the pronunciation that reached Latin America; while in others (areas of ceceo), it is a sound that resembles an English th (or Castilian c/z) more than an English s (or Latin American s), but that is coronal instead of interdental (in an English th or a Castilian c/z there is no sibilance, while the sound in "ceceante" Andalusian is a sibilant, that is, it is actually an unusual kind of s rather than a true th). Another widespread misconception is to confuse the use of this particular kind of th-like sibilant in those varieties of Andalusian with lisping.
All in all, a word like zapatos (shoes) is pronounced (more or less) like thah-PAH-toss in Castilian, like sah-PAH-toss in American Spanish, and (more or less) like thah-PAH-toth in some varieties of Andalusian. Likewise, the words caza (hunting) and casa (house) are perfectly distinguishable in speech for Castilian speakers, which pronounce them (more or less) like KAH-thah and KAH-sah, respectively; while those words are homophonous for the speakers of other dialects, who only differentiate them in writing, both sounding KAH-sah for Latin Americans, and both (more or less) KAH-thah for some Andalusians.
In short, the phonological phenomena just described for Spanish have nothing to do with true lisping and calling them "lisping" or saying that the speakers of certain varieties of Spanish "speak with a lisp" is an improper use of the term which leads to misunderstanding and falsehood. True lisping (i.e. the speech disorder that prevents a person from being able to pronounce [s]) is just as (un)common among speakers of (any variety of) Spanish as it is among speakers of other languages such as English.