Towards a Standard Orthography – Part 2

Part 2

In part 1 of this post we discussed the nature of orthography and orthographic standardization, using the history of French orthography as an illustration. In this post, we will summarize what has been accomplished to date in Cree country with regard to orthography, summarizing the development of three types of Cree orthographies. The first part of this post will summarize the evolution of a Cree orthography based on the French spelling conventions, which culminated in a standard for the a number of dialects spoken in eastern Quebec and Labrador. We will then summarize the development of the English-based orthography. Lastly, the advent of Cree-based orthographies, both syllabic and alphabetic, will be briefly summarized. In part 3 we will expound on the latter and propose some improvements to the consistency of their use by presenting a principled approach to standardization.

French-based Orthography

Page one of L’oraison dominicale, a collection of prayers translated by the Jesuit Massé. Published in 1632, this was the first substantial attempt to represent the Cree language in writing, here using a French-based orthography.

The earliest extant example of an attempt to write anything in Cree goes back to Samuel de Champlain’s reports of his voyages in eastern Cree country, roughly the St-Lawrence region around what is now Québec City. In 1632 he notes the name of a headman likely named Mahîhkanâhtikw, which his report notes as “mahigan aticq.” In the same year, we find a more substantial attempt with the publication of a set of prayers, including the Lord’s Prayer, by the Jesuit Énemond Massé, again in an eastern dialect. The first phrase of the prayer, ‘Our Father who art in heaven,’ is translated in Massé’s publication as “Novtavynan ca tayen ouascoupetz.” While the intrusive pe might have erroneously crept in during the printing process, his spelling reflected French orthographic conventions from the early 1600s. For instance, his use of v, which has since become u in modern French, the use of the letter c to represent a [k] when followed by an o or a, and the use of the digraph ov to represent the sound [u], which modern French now spells ou. When the spelling is brought into line with standards that will be discussed in the following post, this phrase becomes a clear Nôhtâwînân kâ ihtâyan waskohc.

Around the same time, the Jesuit Paul Le Jeune provided a first brief description of Cree grammar in his Relations of the mid-1630s. By the late 1600s, Jesuits such as Antoine Silvy and Bonaventure Fabvre would leave us the earliest extant dictionaries of the Cree language. These Cree-French dictionaries, manuscripts that were not published at the time, presented an orthography based on French conventions of the mid-1600s, the biggest change from Massé’s work being the replacement of the v by u, and the use of the ligature ȣ (commonly represented in modern transcriptions by the numeral 8) for the digraph ou. Their works reflect other peculiarities of French orthography, including the occasional use of tto represent the sound [ts] when followed by i, a common feature of modern Canadian French where a word like petit is pronounced [p(ə)tsi]. They would also make use of the digraph ch to represent the sound [ʃ] and tch to more consistently represent [ts]. These works are nonetheless quite impressive, with a high degree of accuracy despite the awkardness of French orthographic conventions. A key feature missing in these works is vowel length distinction proper to Cree, but lacking in French orthography.

By the 1720s, the Jesuit Pierre-Michel Laure compiled a French-Cree dictionary and translations of prayers by working closely with a Cree-speaking woman named Pechabanoukoue (his spelling). His orthography would closely approximate those of the 1600s, except for his replacement of ȣ for the letter u. By the 1760s, the last Jesuit assigned to Cree country, Jean-Baptiste de La Brosse, would start compiling his Cree-Latin dictionary and would also publish religious literature. While still relying on French spelling conventions, which were yet progressing towards modern French conventions, La Brosse does appear to make an effort to update the orthography by making it more consistent. He replaces ch by shtch by tsh, and also attempts to do away with the use of the letter t to represent [ts], spelling it tsh instead. For instance, the earlier spelling of asti was standardized by La Brosse to astshi (a dialectal form of the word for ‘earth,’ askiy).

By the 1800s, the Oblates had taken over for the Jesuits, two of which, Flavien Durocher and Charles Arnaud, would compile dictionaries of their own. The orthographies in these dictionaires closely mirrored the orthographic conventions set by La Brosse in the century prior.

All of the above works were of dialects spoken along the St-Lawrence River and up the Saguenay River. The orthography developed by these missionaries was based on French orthographic conventions. Centuries later this same French-based orthography would heavily influence the orthographic standardization of descendent dialects spoken in roughly the same regions, now referred to as Innu dialects, as well as the Atikamekw dialect. It should be noted, however, that Mashteuiatsh, politically considered an Innu community, has not endorsed this standard orthography now in use in other Innu communities, though it continues to use a French-based orthography.

A modern example of the standard used by eastern dialects taken from Innu-aimun.ca that discusses the history of this orthography, the culmination of the French-based orthography discussed here.

A brief note will suffice regarding other French-based orthographies outside of the region described above. Louis-Philippe Vaillancourt in 1991 published a French-Cree dictionary of the dialect spoken in Eastmain, which is mixed coastal East Cree dialect with features of both the northern and southern dialects. His intended audience were French-speaking, and his orthography reflects that fact. We should also note that the Catholic missionary Albert Lacombe published a dictionary of the Plains Cree dialect in 1874 that used a French-based orthography. His fairly consistent representation of vowels and consonants would influence the development of a Cree-based orthography, which will be discussed below.

English-based Orthography

Since contact with the English-speaking world in the 1600s, employees of the Hudson’s Bay company routinely noted down Cree names and words in their journals. A few, such as Henry Kelsey in the early 1700s, James Isham in the 1740s, and Alexander Mackenzie in the 1780s made early attempts at gathering word lists. These lists, as with most early examples of Cree written by English-speakers, are barely decipherable largely due to the reliance on English spelling conventions. They stand in stark contrast to early documentation by French-speaking authors largely because of the quirky nature of English orthography and a not so insignificant thing called The Great Vowel Shift. For instance, commenting on James Isham’s spellings of Cree words, Pentland noted in 1977 that Isham “spelled nikotwās(ik) ‘six’ as Cutte wash, cut ta wash ick and Coote washick; he wrote Neish cock, neishcook and neishcoock for niskak ‘geese,’ but Nish ko for the singular niska ‘goose.’

The Great Vowel Shift was a change in the way English was pronounced that spread from dialect to dialect beginning in the 1400s in southern English and influencing all dialects by the 1700s. However, English orthography has largely been standardized by the 1500s, resulting in spellings that deviate considerably form modern pronunciation. It also resulted in multiple pronunciations for a single consonant or vowel, further complicating the picture. Take for instance the words bat, bad, bate, bathe, calm, and calamity, a series of six words with six different vowel sounds; [æ], [æ:], [e], [e:], [ɑ:], and [ə], respectively. Consonants are not any better. Consider the digraph gh in the words aghastenough, and daughter, each treated differently with the first two pronounced [g] and [f], respectively, and the third example simply left unpronounced. This inconsistencies caused by the Great Vowel Shift and other changes to the language after the orthography had been largely settled also resulted in words spelled exactly the same, but pronounced differently. Take for instance the words sow, ‘to plant seeds’ and sow ‘a female pig,’ or lead, ‘a type of metal’ and lead ‘to show the way.’ There are countless examples of this type in this language.

It is no surprise then that early attempts by English-speakers to write Cree were quite frankly terrible. However, by the mid-1800s Anglican missionaries such as John Horden and Edwin Watkins made advances in standardizing their English-based orthographies. Though yet imperfect, what they achieved was remarkable considering the obstacles. Watkins would publish a dictionary in 1865 which continues to be used especially in its second edition revised by Faries and Ahenakew and published in 1938. A Cree grammar publish in 1948 by H.E. Hives illustrated well the evolution of this orthography. However, this English-based orthography is not without problems. It occasionally obscures relationships between certain words and also obscures common processes such as vowel-lengthening. And while it represents certain vowels well, it does poorly at representing the distinction between short and long vowels in many instances.

A page from Hives’ 1952 grammar that illustrates the English-based orthography at its zenith.

The problems inherent in using English-based spelling conventions to devise an orthography for the Cree language, and the shortcomings of what has been achieved compelled linguists in the later half of the 1900s to start looking at creating an orthography based on the Cree language’s own sound system.

Cree-based Orthography

Excerpt from Thomas Vincent’s 1886 translation of Pilgrim’s Progress that illustrates a later stage of development of the syllabic orthography.

By the 1840s a new Cree-based orthography emerged that did away with the Latin alphabet and made use of what has come to be term syllabics. This limited set of characters, based on the unlikely marriage of Devanagari and Pitman shorthand, spread like wildfire due to its ease of use. By the early 1860s, the Anglican John Horden and his team would improve the system by removing the Pitman derived finals, replacing them by smaller versions of the Devanagari derived syllabic characters. This change greatly facilitated the use of this spelling system, though for unknown reasons it would only be endorsed by publications of the Anglican church and the people in the communities around James Bay that this denomination serves. Throughout Cree country, from James Bay to the prairies, an impressive number of books would be published in syllabics throughout the later half of the 1900s.

The beauty of syllabics is that it essentially captured all Cree phonemes well, including representing the distinctive short versus long vowels. In fact, those who created the system were perceptive enough to not mark length on the phoneme /ê/, despite it always being a long vowel, as this one has no distinctive length variation in any dialects of our language. What resulted was a beautiful and accurate orthographic representation of our language that could now move on to the important step of standardization. However, despite the superiority of this spelling system, the task has never taken up. As a result there remain many inconsistencies in this orthography in every Cree dialect that uses it. In fact, inconsistencies have only increased in recent decades due to the gradual deterioration of our language and an insistence by speakers to spell phonetically rather than phonemically. The lack of any real literacy and the absence of official endorsement of standardized syllabic orthographies by influential organizations has allowed this happen, making it now difficult for many to learn how to use what was once the easiest spelling method to acquire.

Excerpt from Bloomfield’s 1930 “Sacred Stories of the Sweetgrass Cree” that illustrates well his phoneme-based orthography.

By the early 1900s, the famed linguist Leonard Bloomfield started gathering data on various Algonquian languages, including Cree, and proposed a first reconstruction of the theoretical ancestral language that would be termed Proto-Algonquian. His transcription of our language set the tone for a use of the alphabet that was neither English-, not French-based, but rather based on phonological features of Cree itself. Michelson, another linguist actively investigating Cree in the 1930s, would also produce phonemic and phonetic transcriptions of Cree, but would also collect impressive amounts of syllabic texts written by Cree-speakers themselves.

As far as we can tell, the first to specifically propose in print the standardization of Cree orthography was C. Douglas Ellis. As a linguist with experience investigating Cree dialects around James Bay, he published his “A Proposed Standard Roman Orthography For Cree” in 1970, proposing an orthography that is based on the internal features of Cree itself. This would be followed up by a 1973 publication by the same author and a 1977 publication by David Pentland, another linguist, titled “Nêhiyawasinahikêwin: A Standard Orthography for the Cree Language.” Louis-Philippe Vaillancourt would also publish on the topic after the publication of his French-Cree dictionary of the dialect spoken in Eastmain. Arok Wolvengrey, another linguist working out west on Plains Cree, and Jean Okimāsis would add to these publications with their own “How to Spell it in Cree: The Standard Roman Orthography,” in 2008.

The proposals outlined in these publications have led to a much-improved representation of the Cree language when using the alphabet, one that is neither based on French or English orthographic conventions. However, the degree to which the proposed conventions are applied varies from dialect to dialect. And while the orthography is intended to be largely phonemic, there continues to be some inconsistencies that are revealed only when tested by using language internal processes.

In the final post we will discuss the proposed standard and compare its application in the real world. We hope to illustrate some of the inconsistencies that should be addressed and will propose ways to do so. A standard syllabic orthography will also be discussed as will the merits of considering its use.

Towards a Standard Orthography – Part 1

Part 1
Introduction

In February of 2021 Montreal-based publisher Joseph John launched an online campaign to add Cree to Google’s translation software. His campaign has managed to attract quite a bit media attention, but reactions among Cree people have been mixed, with pertinent questions regarding dialect and orthography being raised by many. 

For Google to even be able to add our language to its translation software it will have to address both issues, something with which even we as Cree people continue to struggle. But let us disregard the question of dialect for the time being and focus on orthography to better understand what Google would require to even be able to add our language to its service. This topic is especially pertinent as the solution for Google would ideally translate into a solution for us.

In this post we will describe the nature of orthography and orthographic standardization. As an example from which perspective and lessons can drawn, we will then provide a brief history of French orthography. In the following post, a summary of what has been accomplished to date in Cree country with regard to orthography will be provided followed by an outline for a principled approach to a standard orthography of our language.

What is orthography?

Orthography is a set of shared conventions for writing a language that allows speakers of that language to communicate using the written word. Without shared conventions, written communicate becomes laborious, or even impossible. It is the need to communicate using the written word that compels us to abide by what we perceive to be shared conventions. After all, why write if no one can read? 

In the absence of shared conventions, an orthography may gradually take form commensurate with the need or desire to communicate. In the beginning, competing conventions naturally arise. The process of addressing these and settling on a shared set of spelling rules is called standardization. For most languages, this process took hundreds of years and did not happen organically. To settle competing conventions, most languages benefitted from learned persons who systematically assessed them against their understanding of the history of that language and its phonemic and morphemic structures. And yet, despite pronouncing themselves in favour of one convention versus another, progress towards a standard orthography could only happen if others endorsed their opinions. Let us look at the development of the French standard orthography to better grasp what is involved.

A Brief History of French Orthography

The earliest example of written French goes back to 842 with the Oath of Strasbourg. The French portion of this multi-lingual document would be hardly recognizable to speakers of contemporary French given the evolution of both the spoken and written language since. In fact, French orthography as we understand it nowadays was largely establish between the 14th and 17th century, hundreds of years after the first evidence of its written form.

The concerted effort undertaken by individuals during those few centuries aimed at establishing a set of conventions based on both phonemic and historical principles. The resulting orthography, not yet its present form, but likely recognizable to most contemporary speakers, is largely based on the pronunciation of Old French with an attempt to remain as close as possible to the language’s Latin origins. However, these pioneers had to wrestle with many issues, including one we wrestle with as Cree-speakers – Old French was not a unified language, but rather a set of dialects. As a result, the spelling of certain words became largely etymological with the consequence being a multitude of silent letters. For instance, the French word for ‘time,’ pronounced /tã/, would become temps, to represent its Latin origin tempus rather than its Old French origin tens (whence English tense). Additionally, letters that had been pronounced differently in Latin were retained in French in words where the letters had come to represent the same sound, as in the case of the French letters s and c, both pronounced /s/ in the early orthographies. With time, many etymological errors crept in and orthographic reforms became necessary to address the shortcomings of what was a misguided attempt to model French onto Latin, a completely different language in spite of its ancestral relationship with French.

The many errors that were introduced during that period started receiving attention the 1600s with the establishment of the Académie Française in 1635. The progression towards a standard French orthography would continue with further reforms into the 1800s. But these would not be the last. In 1990 a further set of reforms affected another few thousand words, many of them common words. This reform continues to be criticized and even challenged by some, but it is slowly becoming the norm. Even well-established dictionaries were slow to adapt, with the 1990s reforms finally being included in the 2011 edition of the Dictionnaire Larousse.

All told, French has been written for over a thousand years and has really only achieved a semblance of a standard hundreds of years later. Its crawl towards a standard orthography really only picked up the pace in the 1600s, with the establishment of the Académie Française, the pre-eminent council for matters pertaining to the French language. This council of forty members has played a crucial role in the development of a consistent orthography and has published 9 editions of a dictionary deemed official by much of the French-speaking world. Yet, the heavy reliance on Latin continues to make French orthography clumsy, fuelling fringe advocates of even more drastic orthographic reforms who insist on purging French orthography of silent letters and homophonous letters and letter combinations. Despite the heavy influence of Latin, French orthography works well precisely because it is standardized and is endorsed by all French-speakers countries regardless of dialect.

What about Cree?

In contrast, Cree was first committed to writing in the early 1600s, and mostly by non-Cree people until the late 1800s. And although attempts to develop a suitable orthography have been made since the beginning, the task continues to be complicated by dialectal variance, the choice between alphabet and syllabics, and the ever-present influence of a colonial language such as English or French. To make matters worse, ongoing language loss continues to diminish the potential for real literacy in any dialect. The future may seem grim, but there is hope for both us and Google. In our next post we will ignore the noise and review what has been accomplished across Cree country to date and discuss a principled approach to a real standard orthography.