SlopeQ for the BNC: Query Syntax

SlopeQ for the BNC uses the SlopeQ 2 query syntax. The examples below are customized to show how the syntax can be used for searching the data from the British National Corpus the access to which is provided by SlopeQ. For practical reasons the number of examples illustrating each query in this presentation is very limited. However, a link to a page with all the results is given for each query.

Surface queries

This is the simplest type of queries. Words are written in the query box in their plain orthographic form. The results are occurrences of the particular forms submitted in the query. Compare the query and the results:

acknowledge

#	Left	Match	Right
1	Courts and tribunals	acknowledge	that you are entitled to ‘ natural justice ’ , ie that :
2	It may be that at a low level of a graduated test scheme teachers may wish to	acknowledge	a pupil 's recognition of only an equilateral or isosceles triangle as a triangle
3	You can see why he did n't want to	acknowledge	a family connection with someone who was convicted of murder .
4	The fund is voluntarily administered by members of the Trinity House of Leith , whose Master will	acknowledge	all contributions .
5	In this article , I want to	acknowledge	openly a number of tensions that I experience , and which I suspect others may recognise .

Queries of this kind can include a series of terms (positions in the query). The input is a series of words in regular spelling and the results show occurrences of the whole sequence, e.g.:

in vain

#	Left	Match	Right
1	His sister Charlotte , hoping these efforts were not	in vain	, wrote to a friend :
2	they reflect a commitment to national identity that one will seek	in vain	in most parts of the Arab world .
3	the question whether the emancipators had laboured	in vain	is one which should be handled with care .
4	But he hoped	in vain
5	The Roman Emperor Antonine , while he would have looked	in vain	for the wall he built across Scotland 2,000 years before , would have been quite at home in the vast amphitheatre of Celtic Park .

Queries of this kind can be fruitfully used with set phrases, some phrasal verbs and some collocations.

Base form queries

These queries make use of the lexical annotation of the data in the corpus. The results are the occurrences of different forms of the given word. The query is written in triangular brackets as an equation “lemma=”. After the equation mark you put the base form of the word without any space.

The base form of the word is: the infinitive for verbs, the singular for nouns, the positive grade for adjectives. Since verbs are the most inflected words in English we will illustrate this functionality with a verb.

<lemma=decide>

#	Left	Match	Right
1	the constable was required by section 7(4) to inform him that the specimen was to be of blood or urine and that it was for the constable to	decide	which ;
2	The “ yes ” campaign was also supported by the president of the African National Congress ( ANC ) , Nelson Mandela , who , whilst abhorring the notion of the white population	deciding	the political fate of the country , appealed to whites to vote “ yes ” for democratic change .
3	Despite the fact that the state has	decided	to put them to death , many prisoners remain positively patriotic
4	Since little is known specifically about the difficulty factors of reading tabular data it was	decided	to study these in a range of tables .
5	The person	decides	consciously or not to devote time and energy to a particular activity .

This kind of query is a simple way to research the use of the word in all its forms.

In general it is possible combine various functionalities. For example, you can put a surface form and a base form as two terms in one query, e.g.

<lemma=take> advantage

#	Left	Match	Right
1	Betty had	taken advantage	of her weakness , and here she was .
2	recent Budget changes give you genuine cause to wonder whether you are	taking advantage	of the concessions available to you , you should talk to an accountant .
3	Clearly some places , especially a number of fortified towns , were sited to	take advantage	of easily defended positions .
4	The Pension Loan	takes advantage	of the fact that your pension plan can provide a cash lump sum at retirement
5	the waves all round him eased and he	took advantage	of the lull to belt the last thirty yards to the beach .

Such queries are useful with collocations where one of the words always or usually takes just one form.

Operators

Alternative

This operator is represented by the pipe sign “|”. The words separated by the operator are variants of the query term. The results are occurrences of all of them, e.g.:

issue|matter

#	Left	Match	Right
1	It is an excellent way of introducing the	issue	of human rights to school children .
2	The democratic press expressed concern over the threat to freedom of expression by publishing on July 16 a second	issue	of Obshchaya gazeta , a joint paper published during the coup to overcome censorship .
3	There 's still the '	matter	of other women . ’
4	justified the increased deficit “ no	matter	how dangerous for the economy ” because of growing social tensions .
5	but this is a much more serious	matter	if one abandons the assumptions of symmetry and reversibility .

You can input base forms as variants in the query. In that case, occurrences of all forms of all the words will appear in the results, e.g.

<lemma=consist>|<lemma=comprise>

#	Left	Match	Right
1	They usually	comprise	a circle of stake-holes set into shallow trenches with stouter posts located at the doorways
2	the first is the actual rise in new additional funds brought in , while the second	comprises	the increase in existing funds arising from investment performance
3	In brief , what this means is that every individual	comprising	the population of interest should have an equal chance of being selected for the sample .
4	My 1970–3 survey showed that 26% of the TV audience	consisted	of adult males earning less than K450 per annum .
5	It	consists	of spatial visualization , reasoning and experience .
6	to that extent change can be said to	consist	of change in community norms .
7	Under the 1924 Constitution the Republic has an executive President and a National Congress	consisting	of a 120-member Chamber of Deputies and a 22-member Senate ;

As the examples show, the most natural application of such queries is with synonymous words.

It is also possible to use the alternative operator for one of the terms in a longer query. In the example below, the results are occurrences of all the sequences:

benefit|profit|gain from

#	Left	Match	Right
1	They are in a lot of trouble and we have got to	benefit from	it.
2	It was people who Labour claimed would	benefit from	its policies .
3	Both	gain from	being surrounded by the same protective shell
4	But this increase was minuscule when set against the potential	gain from	an improvement in industrial productivity which would make up only half the gap between Britain and its competitors .
5	But the Indians rarely	profit from	the mahogany trees cut from their land .
6	you 'll sell the ice cream you make good	profit from	those and every theatre and every promotional help is done .

Note: you can have more than just two options of the query term, as shown in the example.

Slop factor

This important functionality allows to search for a discontinuous string of words. The query specifies how many words may intervene between the terms of the query. The allowed number of intervening words is set with the slider located below the search box. In this Wiki, the value of the Slop factor is indicated by the following expression: (Slop factor = 1,2,3… etc.).

adopt policy (Slop factor = 2)

#	Left	Match	Right
1	In practice most councils	adopt a policy	of partial delegation , so that routine matters may be dealt with expeditiously by the committees
2	despite the fact that most journals seek to	adopt a consistent policy	and insist on receiving manuscripts which follow a specified system .
3	the internal corporate debate about whether to	adopt a dolphin-safe policy	was ‘ epic , almost theological in tone . ’
4	so that they may	adopt a sensible policy	which we can then follow ?
5	But at the time that the county council had moved to	adopt the policy	, we may be in the working the context of regional policy to be issued by the Secretary of State .

Note: the provided number is the maximum number of intervening words. Strings with fewer words also appear in the results. Strings with no intervening words will also be fetched, but there are none in the corpus in the example query.

Slop factor can be combined with other functionalities, for example alternative:

wait in|on (Slop factor = 1)

#	Left	Match	Right
1	Reassured that our political leaders are both aware of the problem 's growing dimensions and receptive to our rising anxieties , we	wait in	optimistic but realistic anticipation for crime to be at least effectively reduced .
2	But	wait , on	the political horizon there comes a general election .
3	Other nations , especially Japan	wait impatiently in	the wings with their offers of soft loans .
4	I 'll	wait in	… the room , all right ? “
5	Even the Chancellor of Oxford University , Roy Jenkins , has to	wait outside on	the steps of the Clarendon buildings on a cold November morning

Note: punctuation marks count as words for slop factor.

Slop factor with relaxed order

These queries allow intervening words up to the specified number and the query terms may appear in any order. This option is activated by unchecking the “order” box located next to the Slop factor slider (“order” unchecked).

wrong approach (Slop factor = 2) ("order" unchecked)

#	Left	Match	Right
1	On that	approach Halai looks wrong	.
2	Some believe the British	approach is wrong	.
3	Completely the	wrong approach	had been taken by the Government .
4	To ask about details before establishing the general context is to	approach from the wrong	direction .
5	What was	wrong with that approach	?

The relaxed order in pure form is available when the Slop factor is set to 0:

he would ("order" unchecked)

#	Left	Match	Right
1	Surely	he would	.
2	He also said that	he would	press ahead with the controversial plans announced by his predecessor
3	Had she encouraged him	he would	have given more .
4		Would he	be murdering people now in his struggle for existence ?
5	And	would he	be in bed ?

The relaxed order may be combined with other functionalities. For example, in the next query we have slop with relaxed order, there is more than one term and the terms are base forms.

<lemma=abuse> <lemma=right> (Slop factor = 1) ("order" unchecked)

#	Left	Match	Right
1	This form of trespass may also be committed by	abuse of right	of entry .
2	and the defendant subsequently	abuses that right	, then he becomes a trespasser ab initio ( from the moment of entry )
3	( it has always seemed unjust that what was deplored then as a human	rights abuse	was later marketed under the brand name of acid house ) .
4	its leadership replaced and those officers guilty of gross human	rights abuses	brought to trial , had led to the deadlock in the latest round of UN-mediated peace talks
5	says that ‘ all too often , the lack of choices open to the homeless , mean that their	rights are abused	’ .

Notice how each functionality contributes to shaping the set of the results. The use of base forms allows various forms of both words. Slop and relaxed order allow for a variety of constructions in which the queried words interact.

Negation (restricting the results)

This operator excludes specified variants of query terms from the results. Consequently, it must be combined with query types that produce variation in the results. Negation is marked by a pipe sign with an exclamation mark “|!”, which is to be read as “but not”. The example shows how it is used with a base form query. The specified form of the word is excluded from the results:

<lemma=remain>|!remaining

#	Left	Match	Right
1	We	remain	cautious for 1993 , but Bowater remains ready to move wherever the upturn in the business is sighted .
2	Why he did not pay the full amount must	remain	a mystery .
3	and , for a while , the Times of Zambia	remained	the sole daily paper .
4	But Grisedale	remains	a sad place .
5	However , the next owner dismantled the walls in 1685 , since when it has	remained	in a state of decay .

Regex queries

Queries of this type make use of special symbols and quantifiers. Each query is a formula describing a whole set of possible strings of signs (words, sequences of words). The results are occurrences of all predefined strings found in the data.

Wild card and quantifiers

A full stop “.” is a wild card, it stands for any sign.

A plus “+” is a quantifier: the preceding sign can appear one or more times.

An asterisk “*” is another quantifier: the preceding sign can appear zero or more times.

These symbols can be used directly with standard signs, but it is most effective to combine the wild card with one of the quantifiers.

“.+” means that in this part of the query any sign or sequence of signs may appear.

“.*” means that in this part of the query any sign or sequence of signs or nothing may appear.

Note the difference between the quantifiers. If you use the plus, the preceding symbol (which may be any symbol if you use the wild card) needs to appear at least once in each item found. With the asterisk it may not appear at all.

Compare the examples:

develop.*

#	Left	Match	Right
1	for the thesis is that , once	developed	, it is extremely difficult to resist its application .
2	We grow and	develop	as a result of those interactions .
3	So , if there are limits to what we can do , and if the	development	of technology brings costs as well as benefits
4	If we really want to help children	develop	an awareness of which aspects of mathematics are universal
5		Developed	by the Dutch physicist J.R.J. van Asperen de Boer , it employs a camera equipped with an infra-red vidicon that allows an image of a small section of a panel to be transmitted to a television screen .
6	Being a late	developer	, I had been small and slim in comparison to most of my peers , and my appearance had been a large part of my character .

The above query yields words that starting with the letter sequence develop. The very word developis also included. To exclude it from the results, you need to use the other quantifier as below.

develop.+

#	Left	Match	Right
1	There is no shortfall in the East Lothian HP4 supply , nor any difficulty in the availability or	developability	of land .
2	upon the main road frontages , through those which	developed	a system of irregular internal streets and lanes , to the few with some form of organized street plan .
3	CNES	developed	a series of subsidiary companies , entrusted with commercial applications
4	IBM sells the Stratus machines as System/88 , but the new contract includes a cross-licensing agreement giving IBM and Stratus some rights to each other 's patents , suggesting that IBM may have designs on	developing	its own fault-tolerant Unix for the RS/6000 .
5	It was	developments	in sampling theory from statistics which weakened the force of both these assumptions , and quite early on in Britain .

The use of wild card without quantifiers is also possible, but probably less useful for linguistic research. It allows for crossword-like queries which yield word forms of a set number of letters, containing set letters at certain positions, e.g.:

d....e

#	Left	Match	Right
1	The answer to these questions ( particularly the last ) has important ramifications for the raison	d'etre	of the Chinese Wall mechanism .
2	The report says that the methods used by scientists at the Department of the Environment to calculate future acid rain	damage	are ” flawed “ .
3	The author , Susan Hill , was given a large advance to write a sequel to	Daphne	du Maurier 's Rebecca and Emma Tennant is penning a follow-up to Jane Austen 's Pride and Prejudice .
4	Each year in February and March a different	decade	is remembered , with the festival reaching its climax in the year 2000 .
5	She was shrewd enough to see that , if all the saintly characters were killed off , readers might be inclined to	deduce	that ‘ the sure reward of virtue is a fatal accident . ’
6	He kept her waiting , pushing her	desire	to the limits , enjoying the torment , the agony that was surely written all over her face .
7	In the event of my dying before remarriage , I	DEVISE	and BEQUEATH all of my real and personal estate whatsoever and wheresoever not already disposed of as to my freeholds in fee simple and as to my personal estate absolutely to the issue of my union with JACQUELINE MYRTLE MITCHELL
8	McGinlay scored his eleventh goal of the season on Saturday , though he was on the losing side against	Dundee	, and would relish a move to Celtic , no matter their current state .

The elements of regex queries can be combined with other functionalities of the syntax. For example, it is possible to query for a whole group of similar-looking words in all their forms.

<lemma=pain.+>

#	Left	Match	Right
1	It 's also potentially the most disastrous — even Rainey has been caught out by the	painful	highside crash .
2	Not one drop of	paint	hit the wall he was supposed to be painting , but it had done a marvellous job of covering the path and half a garage door .
3	Alcohol tends to be cross-addictive with other mood-altering chemicals such as minor tranquillisers , antidepressants , sleeping tablets and some other prescribed drugs such as Pethidine , Morphia and many other	pain-killers	, anti-histamines and the various illegal drugs such as cannabis , amphetamines , LSD , Mescalin , Ecstasy , magic mushrooms and other hallucinogens , cocaine , heroin and opium .
4	For instance , instead of sending work to the paint shop to be painted , the	painters	were in many cases permanently located at points where work was sufficiently finished .
5	He	painted	scenes in an amateurish fashion from the medieval romance of Sir Degrevaunt on the lower walls of the drawing room , some of which have survived .
6	This was founded in 1696 ; separate music and	painting	academies were established in 1833 .

The above query yields forms of various words starting with pain-. The word pain and its inflected forms like pains are not included because of the quantifier used in the query.

Another possibility is the negation operator combined with the wild card and quantifier, e.g.

vari.+|!variety

#	Left	Match	Right
1	Alternatively , other methods assume some	variability	about a boundary between mastery and non-mastery .
2	This is because	variable	scribal usage is likely to be functional in some way , just as spoken variation is functional ( as suggested in chapter 2 )
3	Thus , ME language states , being so variable , should in principle be suited to the same kind of analysis that we use in present-day social dialectology , and by using	variationist	methods we should be able to explore at least some of the constraints on variation that might have existed in ME .
4	Distinguishing between a multiplicity of variables in accounting for policy	variations	is , of course , a hazardous exercise , but despite the methodological difficulties it seems clear that a simple agency model , with local authorities implementing national policies with little or no discretion , is far from accurate .
5	The knowledge of the likelihood of such modifications and of the probable direction of the changes as suggested by the Hungarian data may be helpful in the explanation of	variance	found in the relationship among developed and , recently , also among developing countries .
6		Varied	programme is envisaged included participation in the Bridge Crosses , a number of services on 3rd & 10th May , schools , Presbytery , church representatives , and Theology & Development students .
7	ME sources , however , also contain variation that may be relevant to non-standard	varieties	and casual styles of speech ; hence , there may be considerable time-depth to these variables also .

The query results do not include examples of the exact form variety, but as can be seen the plural varieties is included.

Grammatical queries

SlopeQ for the BNC makes use of the BNC tagset. We will not present the tagset in full here, but only discuss its application for our search engine. A list of all the tags is available online.

The tags are three-symbol codes, which classify the exact grammatical form of the given word in the corpus. For example, “NN1” marks a singular common noun, “AJC” marks a comparative adjective and “VVG” marks an –ing form of a lexical verb. You can use exact tags like these, but it is even better to make underspecified queries by means of the wild card (and quantifiers). This is possible because the tags form a neatly ordered system. Thus, all tags starting with “N” mark nouns, all tags starting with “V” mark verbs, in particular lexical verbs are marked by tags with “VV” at the beginning, and all tags starting with “AJ” mark adjectives etc.

All grammatical queries have the same formula. They are written in triangular brackets as an equation “pos=” (pos stands for part-of-speech). The tag or regex tag is put immediately after the equation sign. Our first example will use an exact tag, the one for the third person singular present tense of lexical verbs.

<pos=VVZ>

#	Left	Match	Right
1	I 'm sure if ever the occasion	arises	when I want advice on insurance , you 're the first person I 'll come to . ’
2	they can move on to full doctor status and for many students the chance to experience life in another country more than	makes	up for the extra years of study .
3	It also	requires	only a fraction of the fees of other European Universities .
4	The trust has now drawn up detailed plans and	claims	a living museum at Bletchley Park could attract at least 100,000 visitors every year .
5	I do n't suppose anyone really	wants	to see him , do they ?

A formula covering a series of tags can be obtained by using the wild card(s). For example, in order to search for nouns in general you can input <pos=N.+> or <pos=N.*> or <pos=N..>. Of course, these are not identical as regex queries, but because all tags have three symbols, there will be no difference in the results. This use will be illustrated in further examples.

It is possible to combine grammatical search with other functionalities. The following query yields sequences of the word beautiful followed by any noun.

beautiful <pos=N.+>

#	Left	Match	Right
1	A paint-effect wall makes a	beautiful backdrop	, whether you try your hand at sponging or go in for a more adventurous colour-wash finish .
2	HOMES on a north Belfast street have rooms with a	beautiful view	today .
3	Miss Harker removed her bonnet , a	beautiful item	with long blue ribbons , and looked round for somewhere to hang it .
4	the woman could only hope that moving her right away from the influence of the people she went around with into these	beautiful surroundings	might bring her back to herself .
5	And you gave me everything , my	beautiful Maggie	.

This kind of query is very good for researching collocates of a given word that are from a specific grammatical class.

The next query involves the slop factor and a base form. The results are sequences of any form of the word derive followed by a preposition with one intervening word possible.

<lemma=derive> <pos=PRP> (Slop factor = 1)

#	Left	Match	Right
1	More frequently these	derive from	species still found in tropical and semi-tropical habitats
2	The payments pursuant to the discretionary power	derive from	an overseas source and come within Drummond v Collins 6 TC 525 .
3	As a result , it is impossible to	derive egalitarianism in	the Marxist sense from a Biblical foundation .
4	the development of local government reflected economic organization and the political processes which	derived from	it .
5	A cosmopolitan group	derived mainly from	crosses between bush roses and wild species , many noted for their vigour , scent and exuberant flowering .
6	The model is	derived by	the processes of data analysis

It is also possible to apply base form query and grammatical query to a single term. The labels “lemma=” and “pos=” need to be taken in the same pair of brackets then. The following query yields occurrences of all forms of the verb approach, but not of the noun.

<lemma=approach pos=V..>

#	Left	Match	Right
1	As they	approach	a grazing herd of , say , wildebeest , they spread out in line abreast and begin the stalk .
2	she added : ‘ A customer is a customer , I	approach	them in all the same way .
3	When The Art Newspaper	approached	the British Museum they pointed out that they do not possess the Altar of Cybele ,
4	When CDC 's intention became clear , it was	approached	by a variety of potential partners , Ousley says , and ‘ came very close to changing ’ its decision
5	We think now , as Christmas	approaches	, and that elusive fishkeeping present becomes top priority , is the perfect time to get the unbiased recommendations of our experts as to what they would buy , given the chance .

In general, it is possible to add the grammatical specification to any other kind of query term by writing it immediately after the term (without a space). Below we have a regex query for words ending in -fish, but they must be tagged as nouns. In this way we exclude adjectives like selfish.

.+fish<pos=N.*>

#	Left	Match	Right
1	Experienced , mature	Angelfish	can fill two or three Amazon Sword leaves with eggs
2	Over-exploitation has led to a collapse in numbers of	bluefish	tuna , swordfish and cod in the Atlantic .
3	A popular aquarium fish , the range of the Redfin	Butterflyfish	stretches from the tropical Indo-Pacific to South Africa .
4	We have a 10 gallon tank with an undergravel filter containing two common goldfish , one fancy	goldfish	and a Moor .
5	Ian Lucas spots some new opportunities with mid-price	catfish	.
6	Echinoderms , like	starfish	, urchins and sea-cucumbers can be found here .

Tools

Table of Contents