Log in / Register
Home arrow Language & Literature arrow Extending the Scope of Construction Grammar

Epistolary formulae and constructional diffusion

Epistolary formulae as constructions

The language of private letters, and of our seventeenth-century private letters in particular (Rutten and van der Wal 2012; Rutten and van der Wal 2013) is characterized by a large number of so-called epistolary formulae, that is recurrent expressions mainly or exclusively found in letters. In this respect, the letters in our corpus do not differ from, for example, English, German, Scottish or Finnish correspondence of the Early and Late Modern period (Austin 1973; Tieken-Boon van Ostade 1999; Nevala 2004; Dossena 2007; Elspafi 2012; Laitinen and Nordlund 2012). All these letters bear witness to a pervasive pan-European tradition of letter writing dating back to the medieval rhetorical art of letter writing, the ars dictaminis, and Latin and French models for business and legal writings, the ars notaria (see Nevalainen 2001; Poster and Mitchell 2007, and the references there).

Epistolary formulae may fulfil various functions and are used throughout letters, though most formulae occur in the opening and closing passages. Some formulae cover intersubjective domains such as greeting and health. Example

(10), taken from subcorpus 1, presents an opening formula widely used in seventeenth-century letters. Example (11), taken from subcorpus 2, includes the

formulaic expression fris en gesond ‘fresh and healthy, in good health’, repeatedly found in eighteenth-century letters.

(10) Een Vryendelijcke groetenysse sy gescheuen aen

a friendly greeting be written to

‘a friendly greeting be written to’

(11) dese diend om UEd te Laten Weten dat ik Nog fris en gesond zyn

this serves to you to let know that I still fresh and healthy am

‘this [letter] serves to let you know that I am still fresh and healthy/

in good health’

Other formulae do not fulfil a concrete intersubjective function, but rather make the structure of the discourse explicit, for instance by announcing a new topic (12) or by preparing the reader to the closure (13). Both (12) and (13) are taken from subcorpus 2.

(12) Verder heb ik u, myn hertie lief te melden, dat

further have I you my heart love to inform that

‘furthermore, I must inform you, my dear love, that’

(13) Verders niets sonderling meer te melden hebbende als

further nothing special more to inform having than

‘furthermore, having/I have nothing special/other to say but/than’

A number of formulae have relative (pronominal) adverbs. In the remainder of section 6, we will focus on two such formulae. Example (14) presents the address of a seventeenth-century letter, taken from subcorpus 1, which includes the formula waer op Commandeert ‘lit. where on commands, on which commands’. This is a frequently used formula identifying the ship on which the addressee should be found.

(14) Aen reijnier witte Chirurgijn op ’t slands schip de spiegel, to Reijnier Witte barber surgeon on the country’s ship De Spiegel waer op Commandeert den hr viceadmirael michiel de ruijter

REL on commands the Mr vice-admiral Michiel de Ruijter

To Reijnier Witte, barber surgeon on the country’s ship De Spiegel, on which the vice-admiral Mr Michiel de Ruijter commands’

Example (15) gives the opening of a letter from subcorpus 1. It contains the formulaic expression daer uijt dat ick versta ‘lit. there from that I understand, from which I understand’, which is often used to create the transition from reference to a previous letter, to the contents of that letter.

(15) Gonstige vrient ick hebbe ul aengenamen wel ontfang daer uijt dat Kind friend I have your pleasant well received REL from that

ick versta dat ul getrout is

I understand that you married is

‘Kind friend, I have well received your pleasant [letter], from which I understand that you have married’

Epistolary formulae such as (10-15) are similar to constructions in the sense of construction grammar (cf. e.g. Croft 2001; Goldberg 1995, 2006). The following well-known table of the syntax-lexicon continuum, consisting of almost entirely schematic constructions on the one hand to fully lexicalized constructions on the other hand, is based on Croft and Cruse (2004: 255).

Table 4: The syntax-lexicon continuum in construction grammar (Croft and Cruse 2004: 255)

Construction type

Traditional name


Complex and (mostly) schematic


[Sbjbe-Tns Verb-Pp by Obl]

Complex, substantive verb

subcategorisation frame

[Sbj consume Obj]

Complex and (mostly) substantive


[k/ck-Tns the bucket]

Complex but bound


[Noun-s], [Verb-Tns]

Atomic and schematic

syntactic category

[Dem], [Adj]

Atomic and substantive


[th/'s], [green]

An important criterion employed in constructionist approaches to grammar is that of non-compositionality, i.e. the observation that constructions may carry meanings independent of the words they consist of (e.g. Goldberg 1995: 1, 2006: 5). The epistolary formulae discussed here, however, appear to be compositional. The meaning of the utterances is easily derived from the individual words making up the utterance. It has also been argued that utterances may be stored as constructions “as long as they occur with sufficient frequency” (Goldberg 2006: 5), resulting in the conventionalization of utterances as idioms (Bybee 2006). The epistolary formulae discussed in the foregoing are all highly frequent sequences in our letter corpora, which in fact leads us to consider them as formulaic. The opening formula in (10), for example, occurs 41 times in subcorpus 1, and never more than once in a letter, which implies that 41/219 or approximately one in five letters begin with this formula.

Most epistolary formulae consist of a string of words, so they are complex, not atomic. They also contain quite some lexical material. Therefore, we may characterize many epistolary formulae as complex constructions with at least one and mostly several substantive elements. A constructionist description of the epistolary formula in (14) would need to take into account not only the grammatical and semantic properties of the formula, but also co- and contextual information (cf. Bergs 2010). If we term the formula the on which commands- construction, we can specify its basic syntactic form as [Rel op commandeert Subject], meaning that it consists of a relativizer, i.e. a d- or w-form, the lexical- ized string op commandeert ‘on commands’, and a subject identifying the actor who is commanding. These elements can be specified even further. The first variable element, the relativizer, is either daer or waer (in the most common seventeenth-century spelling), that is either the historic form or the incoming variant. The second variable element, the subject, is always a proper noun, i.e. the name of a captain. Moreover, the entire address as presented in (14) is composed in accordance with certain discourse rules. As to the co-text, the on which commands-construction, as a relative clause, is always attached to a noun, i.e. the name of a ship. In (14), this is the ship De Spiegel (‘The Mirror’). From an even wider perspective, the description of the construction needs to account for the fact that it occurs in the address on a letter, which means that letter writers have control of very specific generic rules.

Describing the on which commands-construction as such enables us to cover many examples that we find in our corpus. Diving further into the data, however, we establish that the formula exhibits considerably more variation. In (16-18), taken from subcorpus 1, the meaning COMMAND is not lexicalized by the finite verb form commandeert, but by a noun specifying the social role of the person who commands instead. In addition, the subject of the relative clause (Sijmoen Kerseboom) occurs immediately following the relativizer in (18). Finally, the use of the formula is extended to other parts of the letter. While most tokens occur in addresses, example (18) is taken from the body of a letter, where the writer indicates the ship with which he has sent his previous letters.

  • (16) scheip prinssesse louvijsse daer Commedeur op is Aert Jansz van es ship Prinsesse Louvijsse REL commander on is Aert Jansz van Es ‘ship Prinsesse Louvijsse, on which Aert Jansz van Es is commander’
  • (17) het schip rotterdam daer cappetein op is de manachtiggen lendert the ship Rotterdam REL captain on is the manly Lendert ariiense aeswandt

Ariiense Aeswandt

‘the ship Rotterdam, on which the manly Lendert Ariiense Aeswandt is the captain’

(18) het Schip daer Sijmoen kerseboom Schipper op is

the ship REL Sijmon Kerseboom shipmaster on is

‘the ship on which Sijmon Kerseboom is shipmaster’

These examples are not accounted for by the description given above, which means that we have to assume another construction. Semantically and pragmatically, however, examples (16-18) clearly resemble (14). They are likely part of the same family of constructions, as they fulfil exactly the same function as (14). A constructionist description of examples (16-18), which we consider examples of the where the officer is-construction, specifies the basic form of the construction as [Rel (Subject) commandeur/ schipper/ kapitein op is (Subject)]. Note that in this description the lexicalized part is variable, as it identifies three possible nouns referring to the function of the officer that is in command (commander, shipmaster, captain). As shown in (18), the subject position, as before a proper noun, is variable. The first slot is similar to the first slot of the on which commands-construction, i.e. a d- or w-relativizer. As before, the relative clause always follows the name of a ship. Contrary to the on which commands- construction, the where the officer is-construction not only occurs in addresses, but also in the body of letters.

As explained in the foregoing, we have two formulae, the on which com- mands-construction and the where the officer is-construction, the basic meaning of which is to determine the person commanding a certain ship. From a constructionist perspective, it seems more than likely that these two constructions are connected, in the sense that their resemblance should be accounted for in a representation of letter writers’ linguistic resources. Apart from the similar basic meaning, there is the empirical fact that most examples of both constructions occur in letters’ addresses. We, therefore, assume an abstract meaning COMMAND, which can be expressed by tokens of two constructions. Both constructions are much more abstract than expected on a first inspection of the data. They are more schematic than the typical idiom that they appeared to be, allowing for variation in wording, syntax and position in the wider discourse. Moreover, the single lexical item that the two subtypes share is the preposition op ‘on’. In sum, when discussing (14) we stated that the seemingly idiomatic on which commands-construction is often found in letters’ addresses, but it rather appears that the meaning COMMAND may be lexicalized by a verb (comman- deert) or by a noun and the verb is (commandeur is, schipper is, kapitein is). Taking into account examples such as (16-18) forces us to assume a more schematic representation of the epistolary formula at hand, particularly because the meaning COMMAND can be lexicalized in various ways. This schematization co-occurs with increased variability in other areas: the subject can be any high officer, not just a captain; the subject position is variable; the position of the preposition is also variable; the formula is not only used in addresses, but may occasionally turn up in the body of a letter.

Building on Bergs (2010), who proposes an onomasisological analysis of English future constructions as opposed to the predominantly semasiological approach in constructionist work, we consider the two constructions as members of a family of constructions expressing the meaning COMMAND. Bergs (2010) suggests that grouping (sub)constructions from an onomasiological perspective may be visualized by listing a number of semantically related constructions within a box. Note that the boxes and lines-visualization often employed in construction grammar cannot be used here, as these lines usually denote inheritance relationships based on form, whereas we focus on meaning. Figure 2 is merely meant as an illustration of what such an onomasiological approach to constructional representation entails.

COMMAND constructions

Figure 2: COMMAND constructions

A similar line of reasoning applies to example (15) above. The subcorpora contain many similar examples, and there appear to be clear syntactic, lexical and discourse patterns. The recurrent use of relative clauses with the preposition uit ‘from’ and a form of the verb verstaan ‘understand’ gives the impression of formulaic language, especially because these relative clauses are mostly tied to a noun such as brief ‘letter’ and occur in the opening of a letter. Note that the relative clause is tied to the meaning LETTER, not necessarily to the noun brief ‘letter’. In (15), for example, the formula follows ul aengenamen ‘your pleasant’, where the noun brief ‘letter' is omitted. If we term the epistolary formula in (15) the from which I understand-construction, we could describe its basic form as [Rel uit ik verstaan-Tns]. As before, both d- and w-relativizers can occupy the first slot. Co-textually, the construction follows the meaning LETTER, and in the wider discourse it is important that it appears in openings. Again, a constructionist description would need to take into account this generic pattern, too.

Also in this case, however, we have to assume a more schematic representation, which we call the UNDERSTAND construction. Passing by concrete examples for reasons of brevity, we confine ourselves to the following observations. The first person subject may be omitted, as was common in letter-writing and diary

style. There are also a few examples with a second person subject. A finite verb is also often omitted, writers elliptically restricting themselves to the participle verstaan ‘understood’. Occasionally, writers use a subclause as in (15), inserting the subordinating conjunction dat ‘that’. Furthermore, the verb verstaan may be replaced by similar verbs such as vernemen ‘learn, understand’ and zien ‘see, understand’. This lexical variation mainly occurs in the eighteenth century, where moreover some examples are found in the body of letters. In sum, all the slots are variable, including the generic pattern, only the preposition uit appearing in all examples of the UNDERSTAND construction.

We have elaborated on the constructional representation of these two epistolary formulae for a number of reasons. Epistolary formulae appear to be complex and (mostly) substantive constructions at first sight. The same lexical material occurs over and over again, and almost always in the same place in the structure of the discourse. The greeting formula in (10), for example, viz. Een Vryendelijcke groetenysse sy gescheuen aen ‘a friendly greeting be written to’, occurs 41 times in subcorpus 1, and is always the first line of a letter. Similarly, examples such as waer op Commandeert ‘on which commands' (15) are fairly frequent (e.g. 25 tokens in subcorpus 1), and are largely restricted to addresses. Therefore, these formulae show great resemblance to conversational routines and idioms. Actual language use displays too much variation, however, for an interpretation as fixed idiomatic expressions to be plausible.

Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
Business & Finance
Computer Science
Language & Literature
Political science