What is notable about these 14 sets of words?

accost commit launch deeded
adhere heckle silent versed
ampere arrive angers theses
chance ascent tester stored
chaste ascend tinder seeded
cohere heckle silent versed
debase mucked refile stress
demure riffle silent versed
endure shelve riling nested
forego recess gossip needed
herein advice resins stones
rebate locket career tested
recast costar pirate esters
remade astute surest rested
accost commit launch deededaccolade communed stitched
adhere heckle silent versedadhesive hecklers relented
ampere arrive angers thesesamaranth perigees reverses
chance ascent tester storedchastest ancestor centered
chaste ascend tinder seededchastise ascended tendered
cohere heckle silent versedcohesive hecklers relented
debase mucked refile stressdemurest backfire seedless
demure riffle silent versedderisive mufflers relented
endure shelve riling nestedenshrine duellist revenged
forego recess gossip neededforegone recessed gossiped
herein advice resins stonesheadrest revision incenses
rebate locket career testedrelocate backrest teetered
recast costar pirate estersrecopies castrate starters
remade astute surest restedreassure maturest detested

If each six-letter word is divided into three pairs of letters, then combining the corresponding pairs within each row will produce three new eight-letter words.

ACcoST COmmIT LAunCH DEedEDACCOLADE communed STITCHED

Note that this also works in the opposite direction: if we divide all eight-letter words into four pairs of letters, and combine the corresponding pairs.

ACcoLAde COmmUNed STitCHedACCOST commit LAUNCH deeded

These recouplings work as long as we can divide the $$m \in \mathbb{N}_0$$ words from each sequence into $$n \in \mathbb{N}_0$$ groups that have the same length, and combine the $$m$$ corresponding groups to form $$n$$ new words. Below you see for example how $$m =  2$$ ten-letter words can be divided into $$n = 5$$ groups of two letters, that we can recouple into $$n = 5$$ new four-letter words.

CAlaMItoUS NEwsCAstERCANE laws MICA tost USER
INdiVIduAL CHalLEngESINCH dial VILE dung ALES
LOcoMOtiVE INveSTmeNTLOIN cove MOST time VENT
MAloDOroUS THreESomESMATH lore DOES room USES
REcuPEraTE INteSTinESREIN cute PEST rain TEES

The words in the given sequence do not need to have the same length, as long as we can divide each word into $$n \in \mathbb{N}_0$$ groups that have the same length.

ABatTOir OUTmasTERingABOUT atmas TOTER iring
ABbaCIes HORnedNESsesABHOR baned CINES esses
ABstRUse AFTereFFEctsABAFT stere RUFFE sects
ACtiNIas HOOdedNESsesACHOO tided NINES asses
ACtiVEly REStreTCHingACRES titre VETCH lying
AGlyCOne MASticATIonsAGMAS lytic COATI neons

Assignment

If the word passed to the function divide or at least one of the words passed to the function recouple cannot be divided into $$n$$ groups that have the same length, an AssertionError must be raised with the message invalid division.

Example

>>> divide('accost', 3)
['ac', 'co', 'st']
>>> divide('COMMIT', 3)
['CO', 'MM', 'IT']
>>> divide('accolade', 4)
['ac', 'co', 'la', 'de']
>>> divide('COMMUNED', 4)
['CO', 'MM', 'UN', 'ED']
>>> divide('programming', 5)
Traceback (most recent call last):
AssertionError: invalid division

>>> recouple(['ACcoST', 'COmmIT', 'LAunCH', 'DEedED'], 3)
['ACCOLADE', 'communed', 'STITCHED']
>>> recouple(('ACCOLADE', 'communed', 'STITCHED'), 4)
['ACcoST', 'COmmIT', 'LAunCH', 'DEedED']
>>> recouple('cane laws mica tost user'.split(), 2)
['calamitous', 'newscaster']
>>> recouple('INDIVIDUAL CHALLENGES'.split(), 5)
['INCH', 'DIAL', 'VILE', 'DUNG', 'ALES']
>>> recouple('programming computer games'.split(), 5)
Traceback (most recent call last):
AssertionError: invalid division