Get a PDF of the 2024 edition here

2008 Edition of the
Online Guide to the Engima

Preface to the Guide
History of the NPL
Membership Information
Sharing the Puzzles
      Sharing the fun
      Combining Talents
Editorial Verses
      Acrostical Enigma
      Apt. . .
      Baltimore Deletion
      Baltimore Transdeletion
      Bigram. . .
      Brookline Letter-Change
      Change of Heart
      Combination Padlock
      Diastichal Enigma
      Enigmatic Rebus
      False derivative
      Group flat
      Head-to-Tail Shift
      Heart Transplant
      Letter Bank
      Letter Change
      Letter Shift
      Order Takeout
      Overloaded. . .
      Phonetic. . .
      Phrase Shift
      Picture. . .
      Progressive. . .
      Redro takeout
      Repeated-Letter Change
      Repeated-Letter Deletion
      Reversed. . .
      Sound Change
      Sound Shift
      Telestichal Enigma
      Terminal Deletion
      Trigram. . .
      Welded. . .
      Word Deletion
      Word Substitution
      Solving the Rebus
      Browse the Flat Pages
Introduction to Forms
      From A to O
      From P to Z
      Form Modifiers
      Constructing Medium Crypts
      Solving Cryptograms
      Other Solving Approaches
      Solving Cryptics
      Composing Cryptics
Reference Books
      Non-Guide Flats
      Non-Guide Forms
      Non-Guide Extras
      Where to Find It
      Form Notation
      Italian Picture Puzzles
      Abbreviated Guide to Flats
      Mobile Guide to Flats

© Copyright 2013 by
the National Puzzlers' League
 Solving and Composing the Rebus and Rebade 
By Treesong

Of all Enigma puzzles, the rebus and anagram offer the most play to the composer’s imagination. By the same token, they are the two types most subject to abuse. I’ll discuss some things you may encounter in rebi. (I use rebi, a traditional NPL joke plural, throughout this article. The joke plural of suber - a rebus in reverse - is iber.) Some information may not be completely clear until you’ve been solving for a few months; none will be clear until you have read the rebus and rebade sections in types of flats. Don’t assume I’m endorsing every rebic practice described here!

In most of the following examples, for clarity, the rubric’s reading is given in parentheses after the answer, with implied words supplied in brackets where needed.

The simplest sort of rebus refers solely to the letters and other symbols in the rubric and to their positions. Examples are THALE = in the altogether (in THE, AL together), CI C = catatonic (suber: CI not at a C), and WY = right of way ([to the] right of W [is] a Y). Interjections sometimes spice things up: BALD = ballooned (BAL; lo! one D). They can be overused: P, for example, is a rather silly rubric for hoopla (ho! O! P, la!). You’ll have to judge for yourself whether a rebus is better with, for example, an O left in the rubric or taken out and treated as an interjection in the reading.

Even very simple rebi can raise questions. What is the best rubric for finish in front? Probably most people would say H IFS (F in IS; H [is] in front), where the H is read before (in front of) IFS. But many argue that IFS H is proper, where in front means “farthest along in the direction of advance.” Other prepositions share this ambiguity, and you may see rubrics of either sort. The use of time adverbs and prepositions for spatial relations, as in BT = beret (B ere T), is accepted rebus practice.

Most letters have names: be, bee, ce, cee, aitch, ar, zed, izzard, and so on. The two-letter names are in NI3, not 11C, but they are so commonly used and easy to remember that they are not tagged in readings. (One oddity: ze is not MW, but feel free to use it anyway.)

One phonetic technique became a cliché: the equivalence of “see” and “C”, especially “I see” for -ic word endings. A verse by Hudu lamented an eternal verity of puzzling: “Whenever a rebus is labeled phonetic / There’s always a C in the answer.” This is less of a problem these days; still, if a C is the only phonetic aspect of your rebus, as in O = caring (see a ring), consider leaving the C in the rubric and not making the rebus phonetic at all.

Rebus grammar can be quite condensed, often omitting forms of “to be” and other words. Word order, too, can be quite free, as in O = iodoform (I O do form - for “I do form O”) and H = T for this (T, H is-for “H is T”). Ambiguity is possible: grafter could have the rubric GR (G; R after [it]) or RG (G, R after-for “G, after R”). The latter version is forced but has been used to good effect, as in 0 Y bR = a naughty little brat (a naught; Y; little B, R at). Inversions like these are all right, but readings must be clearly grammatical. For example, L = penal (pen a L) is illegal; the reading would have to be pen an L.

A reading may tell how the rubric was made: D = adipose (a D I pose) or tell the solver how to do it: MYEC = appendectomy (append EC to MY). It may also describe the situation: D = dissolution (D is solution), A = read a paperback of it (read A; paper [in] back of it). One may even have the rubric letters speak: P = imp (I’m P), or be spoken to: GHT = bethought (be thou GHT). Crossing-out is a fertile technique, as in O P = dessertspoon (suber: no O; P stressed) and I = decaffeinates (suber: set an I effaced).

Anything not near the middle of the space above the verse may be an indicator of positional words. Thus a B at the right of the column could be Brighton (B, right, on [= on top of the verse]). XXX at the left could be tench I export (ten, chi, ex [to] port). AB under the verse could be strike a low blow (strike A low; B low). Abbreviations of directions (N, S, E, W, L, R, NE, etc.) have grown common, as in CM/PH [at the right side of the column] = ectomorph (E., C to M or PH). Usually E, for example, implies position at the right of the column, but sometimes it just means position at the right-east-end of the rubric. For example, MYR at the top of the rubric area is a phonetic rebus for wisenheimer (Y’s in high M, E. R).

 Enigmatic and Nonenigmatic Rebi

Enigmatic rebi offer a great variety of ideas, and only experience can give you a feel for the possibilities. “Enigmatic” is not a label for a particular type of rebus; it is more a red flag, warning the solver that something tricky is going on. Roughly speaking, an enigmatic rebus is one whose solution depends on more than the rubric characters, their placement, the situation, or obvious modifications to the puzzle text.

I will mention some rebus types that are ordinarily enigmatic. Note: the terminology is mine, for discussion purposes, and is never used in The Enigma.

In meaning rebi, elements in the rubric are used for their meanings rather than as strings of letters and numbers. In one common type, letters are used as symbols or abbreviations rather than simply as written representations of sounds. An example is this rebus about a child star who can’t play matinees: mm Θ ms = afternoons, the tad is played by male midgets (after “noons” [m = meridies, the abbreviation for noon in 11C], theta displayed by “male” [m = abbreviation for male], midget S), where each m is more than just a letter m. Another is BASiS = basilicons (B, A, silicon, S). More subtle is 7R = neutral Pharisee (neutral pH, ar I see), where 7 represents more than just 6 + 1 (see pH in 11C).

To restrict the enigmatic label to rebi that really need it, the following familiar symbols are treated as standard, not enigmatic: chess pieces and moves (P, Kt, N, B, R, Q, K, 0-0), cards (A, K, Q, J), the basic Roman numerals (MDCLXVI), and postal codes (states, provinces, etc.) in the 11C abbreviations section.

Nonalphabetic characters generally do not make a rebus enigmatic; for one thing, the solver needs no prompting to realize that they are symbols. Unfamiliar symbols not in the “Signs and Symbols” sections of 11C or NI2 are tagged - for example, “Rubric has non-MW material” - to warn the solver to look elsewhere. But the rebus itself will be labeled enigmatic if the rubric involves the meaning of the symbol rather than its simple verbal equivalent. For example, ·-·-·- is the Morse code for “period.” A simple phonetic rebus like ·-·-·- = periodical (period: I see all) is not enigmatic (and really should be done with the simple rubric .). But ·-·-·- = codetermination (code “termination”) is enigmatic, since it is based on the meaning of the period in English punctuation. Another enigmatic example is: C$.10 = catamount (C at amount).

All alphabets are nonenigmatic, including the ones given in 11C under alphabet, manual alphabet, Morse code, and rune. Watch out for Greek letters that look English; P can be rho as well as pe. For example, X = chilies here (chi lies here).

Another type of meaning rebus contains words taken as words: they are used for their meanings, not as groups of letters. For example, Ship = have supper, case the joint (have S uppercase; the joint). Synonym rebi are often just charades in rebus form, but they can be subtler: pen = Aswan Dam (a swan dam - ”dam” in the sense of “mother”; a female swan is called a pen). Another example: Zipangu EB = at last count, rye bread. The reading (at last country, EB [is] read) is dependent on the placement of Zipangu in 11C; it’s the last country in the “Geographical Names” section. Note that words used this way are put in the rubric in their normal form, not in all capitals. Compare JOSIDEKE = sidesplitting joke; here, SIDE and JOKE happen to be words, but the rebus would work exactly the same if they were meaningless strings of letters, so this is not a meaning rebus - and not enigmatic - and the words are not lowercased. Noms are often used in rubrics; since they appear in Enigma both mixed case and all in caps, they can appear in rubrics either way.

Transformational rebi involve alterations to inferred words or puzzle text. An enigmatic example is D     = dauntless aunt (DAUNT less [the letters] AUNT), where DAUNT is transformed by removal of the letters A, U, N, T to produce the final form of the rubric, D. Note: in this subtractive type there should be something to subtract from! ALIS = nominalism (no M in ALISM) is okay, but AL = nominal (no M in AL) is not; there is no reason for that M to be in the reading. The standard nonenigmatic method for subtraction is crossing out: AML = nominal. The ultimate subtraction leaves nothing behind, just an extra blank line or two between title and verse. This could clue stakeouts (S; take out S), gundog (G; undo G), and so on. Naturally, the verse clue to one of these must be good, since it’s doing all the work.

A transformational rebus is sometimes not labeled enigmatic if the “before” of the transformation can be inferred easily from the “after” rubric. A verse that clued forest (for es, T) by changing every S in the verse to T (at in thit parenthetit), or by just retitling the puzzle REBUT, would not be enigmatic. It would be obvious that something had changed, and the nature of that change would not be hard to infer. But a rebus titled SUBER, clueing suer for real sous (SU, ER for RE, also US) is enigmatic, since it looks like a normal suber.

Another nonenigmatic way to show transformations has been to make the rubric a transformed alphabet: ABCDSFGHIJKLMNOP QRSTUVVXYZ = make waves (make W [into] a V [and] E, S).

Transposition (shuffling the rubric letters) is one type of transformation that is not considered enigmatic, whether implicit, as in DIRECT = letters of credit, or explicit, as in GLENLIPS = reformed spelling and CAT = taciturn. A rubric that makes sense, rather than being a jumble of letters, is often a sign of a transposition.

If you can’t find the rubric in the usual place, the puzzle may be a subtle transformational rebus: the period after the puzzle number may be missing (to clue outpoint or pointless), or the author line may be changed (the state FL could be changed to HX to clue fish and listen - F is H and L is ten).

The pictorial rebus regards the rubric (or part of it) as a picture rather than as a group of characters. An example is o o = pieces of eight, where the o’s are seen not as letters but as a picture of a broken-up 8. A similar example is ' ' = split second (the symbol for second - a double quote - is split). Certain pictures are regarded as standard, not enigmatic: O for all sorts of round things (ring, disc, orb); X = cross; and the convention of laying a letter on its back to indicate “sickness,” as in = bill (B ill).

Extrapolation rebi have rubrics referring to words or phrases of which only parts are given: x = deep in the heart of Texas, l = most of all (because L’s make up two thirds of ALL), v = center of gravity. Note that the rubric in each case is lowercase, indicating that it is considered part of a word, rather than a letter with no connection to any particular word. (Some cases are ambiguous.) The rubric in this type is not very helpful, so the verse should clue the answer particularly well. The comment rebus is related, focusing on part of the rubric: BRAN = raisin bran (RA is in BRAN).

In a treasure hunt rebus, part of the puzzle is figuring out where or what the rubric is. (The same problem can also occur in a transformational rebus, as noted above.) For example:

To raise or lower sails that are square,
You’ll be glad that THESE are there.

The solution is: clew lines. Clew is another spelling of clue; the verse is the rubric. Another example was in the June 1990 issue, in which a heart appeared over “June 1990” at the top of page 1. This clued a rebus on an inner page, which had the following tag: “(rubric is earlier in this issue).” The solution was make love on the first date.

The two-level rebus is particularly enigmatic: it is essentially a progressive rebus with the middle step implicit. One example is the phonigmatic K = K* for Kissinger (K is K star, phonetically “K is Kay Starr,” which is a rubric for K is singer). Another phonigmatic example: Sam Bret Hope Stork = four suits (by way of Spade Harte [heart] Diamond Club). Note that, adding to the enigmaticity, one step in the first example is a meaning rebus, as are both in the second example. It’s a general NPL rule that all puzzles are harder than they seem to the composer; the danger of producing an unsolvable puzzle is particularly great for this doubly convoluted type, so be sure to provide good clueing in the verse.

Enigmatic types can be combined, of course. One part of a rubric might be pictorial while another is transformational. A more intimate blend is shown by New Je  ey = out-of-staters (out of [a] state, RS). The fact that everything is lowercase except the natural initial capitals suggests that this is a meaning rebus; the space indicates that it is transformational (subtractive, from New Jersey).

Another possibility, combining a symbolic part with a transformation:  itts urgh = get the lead out (get the lead [= Pb] out). The fact that everything is lowercase suggests that this is a meaning rebus; the spaces suggest that it’s transformational (subtractive, from Pittsburgh), so perhaps the meaningful unit is there only to indicate that a P and B are missing. You may have the solution before you realize that it’s also indicating that the missing P is uppercase and the B lowercase.

 Rebades and Subades

The rebade and subade can use any of the rebus techniques described above. Their unique difficulty is keeping track of which letters of the reading go into which answer words. I find it useful to work with two arrays of dashes, one for the solution words and one for the reading. I work back and forth between the two, replacing corresponding dashes in each with letters as I go. For example, given a rebade with rubric 0 and enumeration (5, 4, *4), if I found that THREE was Erie, then guessed that zero was part of the reading, the diagrams would go from:

- - - - - - - - - - - - -         - - - - -
1 2 3 1 2 3 1 2 3 1 2 3 1         - - - - 
                                  - - - -


- - E - - R - - I - - E -         - - - - -
1 2 3 1 2 3 1 2 3 1 2 3 1         - - - -
                                  E R I E


- - E Z E R O - I - - E -         - Z O - -
1 2 3 1 2 3 1 2 3 1 2 3 1         - E - -
                                  E R I E

And so on until the answer appeared: ozone, Ness, Erie (one zero’s in, see!).

You can also do the array with the solution words going down and the reading across; in that case, the reading line can be eliminated.

- - -
- - -
- - -
- - -
- - E
- - R
- - I
- - E
- - E
O - I
- - E

This works for other types, too. In subades you must remember to reverse when transferring from array to line (or to read the columns bottom-to-top).

I don’t have many rebus/rebade solving hints beyond “Practice!” The following may sometimes help. Given a string of letters, try a variety of prepositions and arrangements to see if something clicks; UMS might be minus, inusem, umats, utomus (Greek mu there), spastum, umans, uthemess, and so on. In solving a suber, try writing the rubric backward to see if it suggests a word. In solving a rebade, try to find a long word and write it alternade-fashion. For example, in solving a five-part subade with R in the rubric, you might learn from the 11C “Signs and Symbols” section that it stood for “ascending node.” You could then write it out like this:


(Remember, it’s a subade, so you would write it out in reverse.) All of the letter sequences (reading down) look likely to occur in words, which is encouraging; and if one of the words clued by the verse contains -nn-, -ge-, -enc-, -dis-, or -oda-, you know how to start filling things in.

Here are two conventions to keep in mind when composing:

Always capitalize rubric letters unless there’s a reason not to. One such reason is to indicate size: for example, Bd could be big-boned (big B, one D) or belittled (be, little D).

Divide up a long rubric to give the solver clues to its structure. One way to do this is to break up the reading into phrases with semicolons and break up the rubric to match. For example: SET OER Y = the Einstein theory (the E in ST; E in the OR; Y).

One last word: be kind to the solver. Unlike other flats, the rebus gives you two places for clueing. If you create a difficult rubric, make the verse very revealing to avoid tiresome obscurity.