The reader of the Romances manuscript soon discovers that the scribe omits or inserts n's and m's - or exchanges one for the other or doubles them - without an apparent system. Another peculiarity is that the sounds /s/ and /ts/, often written ç and tz, are interchanged indiscriminately. Further, untranslatable song syllables, or vocables, appear in most passages, sometimes as i, ya, or iya within a word, often as i, ya, or some other meaningless syllable at the end of a word, producing distortion. A great many other vocables, such as yehua, cahui, and the ubiquitous ohuaya, are free-standing. All have been reproduced in the transcription without any signal to distinguish them from lexical material, just as in the manuscript, and with no assurance that they have been segmented in a way that accurately reflects oral delivery. (For a concordance to vocables, see DICT pp. 729-36.) The following additional points may be noted:

Copyist's errors. Obvious mistakes are abundant, characters misread, whole syllables dropped. Corrective footnotes, where needed, are supplied with the translation.

Roman and italic. The transcription uses two fonts: roman for the heavily penned main text, italic for the more lightly penned jottings of the glossator.

Slash, dash, and bullet. A forward slash, /, found in some passages, evidently marks the boundary between words. A short dash, given in the transcription as a hyphen, -, has the same function, as does the bullet. No doubt it is the copyist responsible for the main text who uses the short dash. The slash, where short and heavily penned, must also belong to the main text, while a thinner, longer slash, often squeezed in as an afterthought, presumably comes from the glossator. The transcription does not distinguish between the two kinds of slash; and wherever the boundary marker is taken to be incorrect, the character string has been closed up, nevertheless preserving the slash, dash, or bullet.

Brackets. Occasional directions to the reader, as well as manuscript line numbers at the beginning of each line, are enclosed in square brackets, [ ], following the usual method of indicating editorial additions.

Paragraph mark. The pilcrow, or paragraph mark, ¶, has been kept throughout, just as it appears in the manuscript (at the beginning of each stanza). It is not an editorial addition.

Stricken and illegible characters. A line through a character, as in a, means that the copyist has stricken it, yet it is still legible. If a character is blotted or otherwise unreadable it appears as the symbol ■.

Abbreviation marks: arc and tilde. The superscript tilde, ˜, and the superscript arc, ̑, occur throughout the manuscript, indicating that one or two unwritten characters follow the marked character, viz., n or m if the marked character is a vowel; ue or, rarely, ui if it is a q; a if it is the p in the combination tlpc (for tlalticpac 'earth', or, as it would seem, the mark converts the entire combination tlpc into the full word tlalticpac). Either symbol written over an n, however, simply converts the n to the familiar Spanish ñ. (Because the arc and the tilde are variously slanted, flattened, or curled in the untidy Romances manuscript, they are often indistinguishable, and the printed transcription, forced to make choices, is therefore provisional in this regard. For more information on the arc and the tilde, see GRAM sec. 4. A third typographic symbol, the overbar, ¯, might have been introduced in cases where the arc or the tilde has been flattened nearly perfectly, but because of the many gradations it is not clear that the copyist intends this. Rarely the arc, where slightly pointed, might be interpreted as a circumflex, ^, instead of an arc, ̑; but again, the scribe fails to make this plain, and the circumflex has not been used in the transcription.)

Abbreviation mark: the vertical stroke. Rarely (at 14:14, 23v:17, 24:16, 27v:19, 40v:16) a vertical stroke, ', appears over a q, converting it to qui (as discussed in Anderson et al. 1976:34).

Diacritics. The arc and the tilde (see above) may be considered diacritics rather than abbreviation marks when appearing over vowels, indicating nasalization, which, as already implied, is anything but consistent in the Romances manuscript.

Calligraphic flourishes. As in other, contemporaneous manuscripts the i is sometimes written with an apostrophe in place of the usual dot. This flourish, even where ill formed and therefore susceptible to being taken as a diacritical mark, has not been carried into the transcription. A more difficult case is the nasalization of the vowel immediately following the ç. The copyist almost always marks such a vowel, using a slanted tilde or an arc, though often the mark strikes the eye as a flourish perhaps inspired by the cedilla under the preceding c (as noted in Karttunen and Lockhart 1980:36). Because nasalization is usually a possibility in such a position, the troublesome strokes have been resolved as either arcs or tildes (both of which are customary for this purpose). Thus the transcription, where it errs, errs on the side of overinclusion, indicating nasalization where a mere flourish might have been all that the scribe intended.

Glossator's numbering. The glossator, or presumably the glossator, has numbered the folios and the stanzas, using arabic, or occasionally roman, numerals (and an odd number 12 appears on the first folio to the right of the title). These have been reproduced, in italic, in the positions where they occur in the manuscript.

In sum, a most vexing paleograph, especially in view of the "diacritics." As reproduced in the transcription, the marked characters are:

The possibility that some of the marks, or their variations, could indicate vowel length or the glottal stop (as in â, á, à, a', etc.) has been ruled out. The main caution is that in some cases, always with vowels, an ornamental flourish might have been the original significance, as noted above. Yet even in those instances, barring an occasional careless slip, nasalization is conceivable. The manuscript does frequently indicate a long vowel by doubling, as in xoochitl 'flower', but seemingly never, or almost never, includes a glottal stop. A rare exception occurs at 21:6-7: tonahaahuiyacan 'let us be pleasured', where the first h represents the stop. The strange-looking mcuhcãtzin 'Dancer' at 1:18 - where the actual phonology calls for mahceeuhcatzin, or mahceeuhcantzin, allowing for a nonsignificant nasal before the tz (and with an h to indicate the glottal stop after the first a) - is here regarded as an anomaly.