From Benjamin Franklin’s Autobiography:
I had begun in 1733 to study languages; I soon made myself so much a master of the French as to be able to read the books with ease. I then undertook the Italian. An acquaintance, who was also learning it, us’d often to tempt me to play chess with him. Finding this took up too much of the time I had to spare for study, I at length refus’d to play any more, unless on this condition, that the victor in every game should have a right to impose a task, either in parts of the grammar to be got by heart, or in translations, etc., which tasks the vanquish’d was to perform upon honour, before our next meeting. As we play’d pretty equally, we thus beat one another into that language. I afterwards with a little painstaking, acquir’d as much of the Spanish as to read their books also.
I have already mention’d that I had only one year’s instruction in a Latin school, and that when very young, after which I neglected that language entirely. But, when I had attained an acquaintance with the French, Italian, and Spanish, I was surpris’d to find, on looking over a Latin Testament, that I understood so much more of that language than I had imagined, which encouraged me to apply myself again to the study of it, and I met with more success, as those preceding languages had greatly smooth’d my way.
From these circumstances, I have thought that there is some inconsistency in our common mode of teaching languages. We are told that it is proper to begin first with the Latin, and, having acquir’d that, it will be more easy to attain those modern languages which are deriv’d from it; and yet we do not begin with the Greek, in order more easily to acquire the Latin. It is true that, if you can clamber and get to the top of a staircase without using the steps, you will more easily gain them in descending; but certainly, if you begin with the lowest you will with more ease ascend to the top; and I would therefore offer it to the consideration of those who superintend the education of our youth, whether, since many of those who begin with the Latin quit the same after spending some years without having made any great proficiency, and what they have learnt becomes almost useless, so that their time has been lost, it would not have been better to have begun with the French, proceeding to the Italian, etc.; for, tho’, after spending the same time, they should quit the study of languages and never arrive at the Latin, they would, however, have acquired another tongue or two, that, being in modern use, might be serviceable to them in common life.
To his reasoning against beginning with Latin, I’d only counter that starting with a language of middling grammatical complexity prepares you to move both up (Greek) and down (any of the romance ones) while starting with one of relative grammatical simplicity can make upward motion more grating. When I taught Latin I found the people who came with a base in French or Italian were more frustrated at the case system and the range of verb tenses/constructions than those who came with no language learning background. It was easier to accept the density of rules as a (frustrating) given when you didn’t have a point for comparison in your pocket.
The editor of the Gutenberg text provides a bonus unsourced footnote from Edward Gibbon – though it seems to me only partially connected to Franklin’s point:
“‘Our seminaries of learning,’ says Gibbon, ‘do not exactly correspond with the precept of a Spartan king, that the child should be instructed in the arts which will be useful to the man; since a finished scholar may emerge from the head of Westminster or Eton, in total ignorance of the business and conversation of English gentlemen in the latter end of the eighteenth century. But these schools may assume the merit of teaching all that they pretend to teach, the Latin and Greek languages.'”