Rare Bigram Drill (gh wk zy xc)
gh wk zy xc pw kn bt lk mn
Every passage contains: gh wk zy xc pw kn bt lk mn
QWERTY layout assumed. Backspace corrects; uncorrected errors count against net WPM.
"gh", "wk", "zy", "xc", "pw", "kn", "bt", "lk", and "mn" don't fail for a mechanical reason the way the other bigram drills in this batch do — no shared finger, no row jump, no cross-hand timing to get wrong. They fail because ordinary typing barely ever produces them, so hands never get the repetition needed to turn any one of them into an automatic, pre-loaded motion the way the Common Bigram Drill's ten pairs long ago became. Each time one shows up, it has to be assembled from its two letters rather than fired off as a chunk the way "th" or "in" would be.
Two different histories explain why these nine pairs are this rare, and they're not the same history. Five of them are leftovers from sounds English stopped making centuries ago; the other four are rare because of where in a word they're allowed to occur, not because of anything historical.
Why This Drill
"kn" in "know" and "knee" once had an audible K. "gh" in "light" and "night" once carried a guttural sound English no longer makes. "bt" in "doubt" and "debt" was added to match Latin spelling even though English never pronounced the B. "mn" in "autumn" and "column", and "lk" in "walk" and "talk", both keep a letter the modern pronunciation quietly drops. The spelling never caught up with the sound, which is a large part of why these five bigrams are rare to begin with — the everyday speech that would otherwise reinforce them disappeared long before the letters did.
The other four are rare for reasons closer to word structure than history. "xc" mostly shows up where the prefix "ex-" meets a root that starts with C — "except", "excite", "exceed" — rather than inside an unbroken word. "pw" turns up almost exclusively at the seam of a closed compound: "upward", "shipwreck", "stopwatch". "zy" is a genuine native ending, just an uncommon one, in words like "cozy", "lazy", and "hazy". "kn" is stranger still, once you notice it: it only ever opens a word in English, never sits mid-word or at the end, which means it gets exactly one chance per word to appear at all.
Because none of these nine pairs gets enough exposure to automate, typing one under speed means briefly building it from its two letters rather than firing it as one learned unit — a small, conscious interruption inside otherwise fluent motion. This drill's only job is to run that specific interruption often enough, on these nine pairs, that each one gets at least some of the automatic treatment everyday pairs receive for free. It sits at the opposite end of this batch from the Hand-Alternation Bigram Drill: that drill's pairs are so common they barely need attention at all; these may never fully stop needing it. The Vowel-Heavy Word Drill is the closest thing on the site to a companion for this one — a different kind of rare pattern, built from letters that are individually common but rarely cluster the way that drill's words force them to.
Where the Silent Letters Went
Old English and early Middle English pronounced the K in "know" and the guttural sound behind "night"'s "gh" out loud. Later, English spelling stopped changing to match its own pronunciation while the pronunciation kept drifting, which is how a word like "walk" ended up carrying an L that most speakers no longer sound. "debt" and "doubt" have a stranger path: their B was inserted well after the words were already established in English, specifically to make the spelling resemble the Latin words they came from, even though no version of English ever spoke that B aloud.
Prefix Seams, Compound Seams, and One Genuine Native Ending
"xc" and "pw" aren't relics of lost pronunciation at all — they're artifacts of where English happens to draw its word boundaries. "except" and "excite" put "ex-" directly against a C; "upward" and "shipwreck" put one word's final P directly against the next word's initial W inside a single closed compound. "zy" is the odd one out among the four, since "cozy" and "lazy" aren't built from two smaller pieces at all — it's simply an uncommon but ordinary way for an English word to end.
Why "kn" Only Gets One Shot Per Word
Every common word carrying "kn" — "know", "knee", "knife", "knight", "knot" — puts it at the very front, because the pattern comes from a specific class of native English words that all happened to begin that way. There's no equivalent English word where the combination shows up in the middle or at the end, which means the practice this drill gives it is the only context that pair will ever actually need.
Frequently Asked Questions
Why does typing one of these pairs feel like more of an interruption than a normal slow patch?
There's no accumulated repetition behind any of the nine to draw on, so your hands are essentially spelling the pair out consciously each time, rather than reaching for a chunk built through repeated use the way a common pair like "th" is.
Are "gh", "kn", "bt", "mn", and "lk" rare for the same reason?
Yes — all five keep a letter English used to pronounce and later stopped, so the spelling is older than the current sound. The everyday speech that would otherwise reinforce these combinations disappeared long before the spelling did.
Why is "kn" only ever at the start of a word?
It's a quirk of which English words happen to carry it — every common example is a native word where the silent K opens the root, and English simply has no words where the combination sits mid-word or at the end.
Is "xc" really as rare as the others, given how common a word like "except" is?
The pair itself is rare even though a couple of its host words are common — "xc" only forms where the prefix "ex-" happens to meet a root starting with C, a narrow, specific junction rather than a combination scattered freely through the language.
Should I expect to ever type these as fast as an ordinary bigram like "th"?
Not to the same degree, since "th" gets constant reinforcement from ordinary text that these nine pairs will never receive in comparable volume. The realistic goal is removing the conscious hesitation, not matching the raw speed of a pair you type hundreds of times a day.
Is this drill worth repeating often, or just once to get familiar?
Occasional repetition, especially whenever one of these pairs trips you up in ordinary typing, does more good than a single pass, since normal text won't hand you enough natural exposure to keep the gains fresh on its own.