Why did Roman authors never feel a need for word spacing?
An interesting question because the Romans certainly accepted the notion of word division, at least until about 100 AD, at which point Romans adopted scriptio continua in imitation of the Greeks, a move that at least one scholar has called a “deplorable regression”. This word division had been marked chiefly via interpuncta, a system probably inherited from the Etruscans, but not spaces between words. So, to rephrase your question a little: if you’re going to have word division, why use dots and not spaces? After all, if “punctuation” or pungere means to puncture or to create a gap, why not insert a literal gap in the text?
Of course, Romans had gaps in their speech. Cicero, for instance, when writing about oratory, often refers to pauses between words (interpuncta verborum De Oratore III.XLVI.181), pauses between ideas (interpuncta argumentorum De Oratore II.XLI.177), and pauses in which to catch one’s breath (interpuncta intervalla, morae respirationesque Orator XVI.53).
But Saenger suggests that Romans’ reliance on reading aloud made indicating these pauses visually with spaces in the text redundant. And while the lack of spaces may have made comprehension difficult, he argues that reading was confined to a small, elite group and thus there was no incentive to make reading easier. Indeed, your ability to decipher a text may have been a marker of your belonging to that elite group.
Reading with your ears, not your eyes
Looking at Latin sources, I can certainly find a lot of evidence of how essential hearing a text was to the understanding of it.
For example, Quintilian insists that a text needs to be heard by his students in order for their eyes to make sense of it:
… et hercule praelectio quae in hoc adhibetur ut facile atque distincte pueri scripta oculis sequantur …
… and indeed reading aloud, which is to be employed, so that the
boys can easily and clearly follow the writing with their eyes …
Quintilian, Institutio Oratoria, 2.5.4
Cicero repeatedly claims that the ear is the best judge of good writing. One example:
… ut qui vellet auribus ipsis quid numerosum
… so that he who will may judge by the ear alone what is rhythmical
Cicero, Orator, LXVII.226
And elsewhere, he writes that a period should be arranged ‘so as to satisfy the ear’ (ut et auris impleat) Cicero, Orator, LXVI.221
Moreover, rearranging the words and therefore the rhythm of a text may keep the sense intact but this is not the point, Cicero argues. It is a practice ‘with which I disagree. “But the words are the same, the meaning is the same” [you say]. That satisfies the mind, not the ears …’
… a quo dissentio. “At eadem verba, eadem sententia”. Animo istud satis est, auribus non satis …
Cicero, Orator, LXIII.215
Lastly, Quintilian on the difficulty of keeping voice and eyes on the same page (as it were) when reading:
… et quod difficillimum est, dividenda intentio animi ut aliud voce aliud oculis agatur
… and most difficult of all, dividing the attention of the mind, so
that the voice does one thing, the eyes another
Quintilian, Institutio Oratoria, 1.1.34
Saenger would argue that it’s only when reading becomes a contemplative, private and silent act that spaces start to appear because this is easier and quicker to comprehend in one’s head. Certainly, many studies have shown that inserting spaces into text improves reading speed, even in languages that do not normally use spaces (including Thai). However, making reading easy and quick was never the aim for Roman authors. Reading wasn’t private, it wasn’t about quiet time or for quickly searching for information, all of which is a very modern understanding of the function of reading.
Instead, the lack of punctuation at time of publication and, indeed, the lack of spaces, meant that all texts needed intense preparation before they could be read with any ease and certainly before they could be read aloud. Parkes notes that reading aloud at first sight would have been “unusual and unexpected”. This is why the praelectio was such a feature of Roman reading and also why texts were marked up by copyists (the notis librariorum Cicero disliked so much), even when interpuncta were common. It also led to confusion, especially once scriptio continua was embraced. Even grammarians could not agree on meaning due to differing opinions about word division. Does Aeneid 2.798 read collectam ex Ilio pubem/a people gathered from Troy or collectam exilio pubem/a people gathered for exile? Is it conspicit ursus/a bear sees or conspicitur sus/a sow is seen?
Therefore, I would argue that Romans did feel a need for “word spacing” insofar as they needed help in discerning the pauses, breaks and gaps between words, ideas and sentences. We have evidence that once the text had been heard, many Romans felt compelled to mark their texts with annotations that approximated what they had heard. However, perhaps because of the reliance on oral/aural cues in reading, it never seemed to occur to them to represent these intervalla visually by inserting literal gaps into text.
Nevertheless, the concept of spacing text was used in one particular context: texts aimed at those learning Latin in the 3rd to 7th centuries AD, a period during which scriptio continua was in vogue. A combination of spaces, columns and/or various punctuation symbols were routinely used to make absolutely clear to the (typically Greek) student where one word ended and another began, or to indicate groups of words that formed a grammatical unit, or to separate lemmata from glosses. Of course, this could use up a lot space and in some cases where space was at a premium, such as in a trilingual text, the words would be run together. But then another less spacious form of word division was employed. It seems a strange irony that Latin gave up word division to mimic Greek scriptio continua but then had to employ word division so that Greeks could learn Latin!