Literacy is a notoriously hard nut to crack for ancient times. To cannibalize some parts of my longer answer on literacy in general in Ancient times here, literacy itself is a complicated concept, that has evolved into more than simply 'being able to read and write', encompassing more fields such as numeracy, the ability to understand complicated texts and the signs and symbols a society uses. Literacy can come in various degrees, and if you define literate very narrow, no-one is 100% literate - do you know every word of the English language? It's a continuum, a person is not either literate or illiterate, but possesses different degrees of literacy. The lowest level of literacy today is often categorized as 'functional illiteracy', that is, those people are able to identify the letters or words, but unable to understand longer sentences or text, while an illiterate person can't even do that.

That being said, I am of the opinion that the threshold of what we call 'functional illiteracy' is much lower in the Ancient world. Today, you can said to be functionally illiterate if you can't understand, say, a newspaper article, but ancient texts that people were confronted with were usually much shorter, and especially on inscriptions or coins highly formulaic, so that you did not need a big corpus of words to understand what was going on. Receipts or invoice might not note much more than the name, the price, and the product, nothing more. While it is true that many slaves would have to have had a basic understanding of the written word would have been enough to 'function' in their profession. Ironically, ancient inscriptions tell us through the errors the stonemason made that the people producing the physical text must often have been illiterate, so the simple reproduction of writing did not necessarily mean that someone was able to read, and in fact, Roman education taught children how to write first, and in many examples, the students seem to not understand what they write even when copying segments of verse, prose or aphorisms - then we have the abundance of abecedaria both in graffiti and in inscriptions, that is simply the alphabet written out. Here people practiced writing, without much understanding of what they wrote, that is, unable to read it in more complicated forms.

Most forms of ancient literacy seems to have moved in this kind of grey area of low-level literacy. What today would be classified as 'functional illiteracy' would have sufficed to function in antiquity. There was no general public education, and many who would have learned to read or write would have done so out of necessity for their profession, for example as a writer for a legion, or as slaves or workers in the grand aristocratic estates.

But not only is literacy a complicated topic, so are slaves. Slaves are difficult to pin down, since they leave so few traces - especially in writing, apart from a few grafitti that were written by slaves, or the names of slaves involved in the production of goods that were marked with their name - and not much more than that. Especially for the latter group, the ability to write their name on a product can not be tied to their ability to read as well. Most of our accounts about slaves come from the elite, and we must always be aware that this clouds our view of them as well. What may have seemed ilitterate to a highly literate Senator trained in the finest points of rhetoric, well-versed in literature and poetry, might still be a decent level. Highly educated slaves and freedman, such as Ciceros famous Tiro are a common literary trope (as well as a target for ridicule as in Petronius' Satyricon, where one such freedman is characterized by his lack of education and reading skills), and the kind of slaves you mention, those tasked with writing down important messages, writing or copying books for their masters or reading to them as Pliny had his slaves do for him would have had to have a high amount of literacy. To quote Harris (p. 256), "There was need everywhere for a number of slaves who were literate in varying degrees, and every man of wealth would require some such." So some of their slaves would be sent to school, and taught to read and write, to some degree necessary for their job.

Varro, in his treatise on agriculture, regarding how to manage your estates and your slaves, says that

Slaves should be neither cowed nor high-spirited. They ought to have men over them who know how to read and write and have some little education, who are dependable and older than the hands whom I have mentioned [Varro, Re Rustica I 17.4]

And, regarding the care for flocks or herds:

In the matter of health there are many rules; but, as I said, the head shepherd keeps these written down in a book, and carries with him the remedies he may need. [Varro, Re Rustica II 2.20]

So here are some example where slaves would need reading or writing to a degree, even though other agricultural writers such as Columella mention that they could do their job just as well when they were illiterate as long as they were smart (Colum. I, 8, 4). However, the level required there was not very high, such low level literacy would have served many well enough to function in a world were text was indeed ubiquitous, but often short, simple and formulaic. A recent trend has been to identify many such 'functional' literacies, or 'specialised' literacies, for people working in certain professions who were only able to truly comprehend or write those texts that they were confronted with in their professional capacities, but others (me included) find it hard to believe that such skills wouldn't have been transferrable (at a low level still, maybe, but a domus is a domus in any text).

But those are the slaves at the top of the hierarchy, the vilici, who oversaw the work of the other slaves, or the amanuenses, their personal secretaries. Apart from those kind of slaves, Cicero, to stay with the example, had a considerably higher number of other slaves, cooks, runners, watchmen, washers (though he might have done some washing himself), watch his estates and various other possessions and so on and so forth. A considerable number of his slaves would have no need of literacy to function in their assigned tasks. In a big city such as Rome, that number would have been high indeed, but in the countryside, it was probably only the overseer who was able to read or write at some level, or the smaller number of slaves employed in specialized professions. Most slaves and their names are known from Rome, and the very fact why we know them (that someone put up an epitaph with their name, often fellow slaves) suggests also that the literacy among slaves there was higher than elsewhere.

Concerning prices, yes, it's true that slaves prices were, apart from the usual effects of supply and demand, also determined by the skill level of the slaves. That could include 'skills' such as being young and fertile (the children of your slaves would become your slaves as well, and thus this was a cheap way to replace your slave stock), but also crafts, and presumably reading and/or writing. The diocletianic edict on maximum prices specifies a maximum price ceiling of 2 times the price for an unskilled slave for a slave that is reasonably skilled, but as we don't really know how the average prices of slaves relate to the effective ceilings, and the edict being itself a highly problematic source, it's hard to say, it's also hard to say at which level reading/writing would have been valued, since the edict only specifies

according to gender and age, and the quality of the skill, it shall be proper to agree the price [as much as 2x as for an unskilled slave].

A passage in Sueton, here quoting a certain Orbilius, suggests that literate slaves brought a higher price, since it was used to advertise them, but that is the only ancient reference to slave prices related to literacy:

"In the days of our forefathers, when anyone's slaves were offered for sale, it was not usual except in special cases to advertise any one of them as litteratus but rather as litterator, implying that he had a smattering of letters, but was not a finished scholar." [Suet. Grammaticis 4].

It certainly enhanced the value of a slave, but high volumes of literate slaves were a thing connected with the Roman conquests of Greek and the East, probably it was easier to train them, as Crassus is said to have done:

nothing compared with the value of his slaves; so many and so capable were the slaves he possessed, — readers, amanuenses, silversmiths, stewards, table-servants; and he himself directed their education, and took part in it himself as a teacher, and, in a word, he thought that the chief duty of the master was to care for his slaves as the living implements of household management. [Plut. Crass. 2, 5-6].

In short, the level of literacy among slaves was probably very low, lower than the average population. Harris has estimated a level of 20% of literacy among the adult population, but while I find it problematic for the population at large, it may be much closer to the level among slaves, even accounting for the larger number of them who would have aquired some level of functional literacy.

Sources and further reading:

- W. Harris, Ancient Literacy, Cambridge (MA) 1989.

- J. Bodel, Inscriptions and Literacy, in: The Oxford Handbook of Roman Epigraphy, Oxford 2015, pp. 745-763.

- B. A'Hear et al., Quantifying Quantitative Literacy: Age Heaping and the History of Human Capital, Journal of Economic History V. 36, 2009, pp. 783-808.

- S. De Brestian, Interrogating the Dead: Funerary inscriptions in Northern Iberia, Montagnac 2008.

- A. Cooley (ed.), Becoming Roman, Writing Latin? Literacy and Epigraphy in the Roman West, JRA Supplement 48, Portsmouth 2002.